record of decision (rod)(signed) - operable unit 2 - …system"). construction of the...
TRANSCRIPT
Record of Decision for Operable Unit 2 - Pine RiverVelsicol Chemical Superfund Site
St. Louis, Michigan
Site Name and Location
The Velsicol Chemical Superfund site is located at 500 Bankston Street, Gratiot County, St.Louis, Michigan and encompasses contaminated portions of the Pine River. There are twooperable units (OU) at the Velsicol Site. OU1 consists of the 52 acre main plant site, the locationof the former chemical manufacturing facility. OU2 consists of contamination in sediments andfish in the St. Louis Impoundment and Pine River.
Statement of Basis and Purpose
This decision document represents the selected final remedial action for OU2, contaminatedsediments in the St. Louis Impoundment of the Pine River. This final remedial action wasdeveloped in accordance with the Comprehensive Environmental Response, Compensation andLiability Act of 1980 (CERCLA), as amended by the Superfund Amendments andReauthorization Act of 1986 (SARA), and to the extent practicable, the National O;l andHazardous Substances Contingency Plan (NCP). The attached summary of RemedialAlternatives identifies the information contained in the administrative record for this site uponwhich the selection of the remedial action is based.
The State of Michigan, Department of Environmental Quality has indicated a willingness toconcur with this Record of Decision. A copy of their concurrence letter will be added to theAdministrative Record upon receipt.
Assessment of the Site
U.S. EPA and the Michigan Department of Environmental Quality (MDEQ) collected sedimentand fish tissue samples from the Pine River that document concentrations of total DDTresulting in unacceptable risk to human health and the environment from fish consumption(humans and fish-eating birds).
Actual or threatened releases of hazardous substances from the site, if not addressed byimplementing the remedial action selected in this Record of Decision, may present an imminentand substantial danger to public health, welfare, or the environment.
Description of the Remedy
The selected remedies are Alternatives 4, 5 and 6 and consist of:* temporary coffer dams or sheet pile;
* removal of all contaminated sediments greater than 5 ppm total DOT usingexcavating equipment or dredging;
* addition of a stabilizing/drying agent to the sediment;* dewalering sediment if necessary* water treatment with discharge to the Pine River;* monitoring of resuspension of sediments during removal;* confirmation sampling of sediments* transportation of sediments via truck to a landfill; and* disposal of sediments in either a RCRA subtitle C or D landfill (depending on
level of contamination in sediments).
Operable Unit (OU) 1 of the site was completed by Velsicol in 1984 under a 1982 ConsentJudgment and consisted of construction of a slurry waU and clay cap over the 52 acre main plantsite. This ROD only addresses contamination in OU2, Pine River sediments.
Sediments contaminated with total DOT are the principal threat at the site. This actionaddresses sediments contaminated with greater than 5 ppm total DOT by removing them fromthe river, dewatering them and disposing them in a landfill.
Statutory Determinations
This final remedy is protective of human health and the environment, complies with Federal andState requirements that are applicable or relevant and appropriate to the remedial action, is costeffective, and utilizes permanent solutions and alternative treatment technologies to themaximum extent practicable. This remedy also satisfies the statutory preference for treatmentas a principal element of the remedy (i.e., reduce the toxicity, mobility, or volume of materialscomprising principal threats through treatment). Although this remedy will not result inhazardous substances remaining in sediments on-site above levels that allow for unlimited useand unrestricted exposure, a five-year review will be required for this remedial action becauseit will take time for contaminant levels in fish to drop to acceptable levels.
ROD Data Certification Checklist
The following information is included in the Decision Summary section of this Record ofDecision. Additional information can be found in the Administrative Record file for this sitewhich is located at EPA's office in Chicago or at the T.H. Cutler Memorial Library in St.Louis, Michigan.
Chemicals of concern (COCs) and their respective concentrations• Baseline risk represented by the COCs
Cleanup levels established for COCs and the basis for the levelsEstimated capital, operation and maintenance (O&M), and total present worth costs;discount rate; and the number of years over which the remedy cost estimates are
projected• Decisive factor(s) that led to selecting the remedy (i.e. . describe how the Selected
Remedy provides the best balance of tradeoffs with respect to the balancing andmodifying criteria).
The following information isjiQj included in the Decision Summary section of this Record ofDecision because it does not apply to this site.
• Current and future lanu and ground-water use assumptions used in the baseline riskassessment and ROD.Land and ground-water use that will be available at the site as a result of the SelectedRemedy.
State Concurrence
The State of Michigan, Depanment of Environmental Quality has indicated a willingness toconcur with this Record of Decision. A copy of their concurrence letter will be added to theAdministrative Record upon receipt.
a/'*/William E. Muno / DateSuperfund Division Director
RECORD OF DECISION SUMMARY
VELSICOL CHEMICAL SITESt. Louis, Gratiot County, Michigan
February, 1999
TABLE OF CONTENTS
RECORD OF DECISIONVHLSICOL CHEMICAL SITE
I. Site Name, Location, and Description.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
II. Site History and Enforcement Activities.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
III. Community Participation.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
IV. Scope and Role of Operable Unit................................................................10
V. Characteristics of Pine River..................................................................... 10Sediment Data Summary.......................................................................... 11Fish Data Summary................................................................................ 13Surface Water Data Summary.................................................................... 15
VI. Current and Potential Future Site and Resource Uses.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
VII. Summary of Site Risks.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Human Health Risks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Ecological Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20Basis for Act ion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
VIII. Remediation Objec t ives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
I.. Description of Alternatives........................................................................21
.. Summary of Comparative Analysis of Alternatives.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
.I. Selected Remedy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
.II. Statutory Determinations.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26Overall Protection of Human Health and the Environment..................................26Compliance with ARARs.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26Cost Effectiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27Utilization of Permanent Solutions and Alternative Treatment TechnologiesTo the Maximum Extent Practicable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28Preference for Treatment as a Principal Element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28Five-Year Review Requirements.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
.III. Documentation of Significant Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
5
IABLES
Table 2.2-3: Concentration, Area, Volume and Mass of Sediments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Table 2.2.1.4-1: Total DOT Concentrations per Depth Interval (1997 Data) . . . . . . . . . . . . . . . . . . . .12
Table 2.2.1.3-2: Total DOT Concentrations in Sediment below St. Louis Dam.... . . . . . . .....13
Table 2.2.2-1: Skin-off File. Carp Samples....................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Table 2.2.2-2: Carp Data Specifics...................................................................... 14
Table 1. Exposure Assumptions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Table 2. Current Risks from Fish Consumption.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Table 3. Dermal Cancer Risks Using 1997 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Table 2.5-1: Volume Break Point Analys i s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
Table 2.5-2: Post Remedial Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
APPENDICES
Appendix A - Figures
Appendix B - Detailed Cost Estimate for Selected Remedy
Appendix C - Responsiveness Summary
Appendix D - Administrative Record Index
I. Site Name, Location, and Description
Site Name: Velsicol Chemical SiteLocation: St. Louis, Gratiot County, MichiganCERCLIS #: MID00072439Site Lead: U.S. EPAType of Site: Former chemical manufacturing facility/contaminated river sedimentsNo.ofOUs: 2
There are two operable units (OU) at the Velsicol Si^ OU1 consists of the approximately52 acre main plant site, the location of the former chemical manufacturing facility. OU2 consistsof contamination in sediments and fish in the lower and middle basins of the St. LouisImpoundment of the Pine River. See Figure 2-1. The Site history contains extensive informationabout OU1 which is provided for background purposes only.
II. Site History and Enforcement Activities
The Velsicol Chemical Site ("Site") is an approximately 52 acre parcel that was onceoccupied by a chemical processing plant and contaminated sediments in the St. LouisImpoundment of the Pine River. See Figure 2-1. The chemical plant operated from 1936through 1978 and manufactured a variety of organic and inorganic chemicals includingpolybrominated biphenyls (PBB), hexabromobenzene (HBB), l,l,l-Trichloro-2,2-bis(p-chlorophenyl) Ethane (DDT), and Tris(2,3-Dibromopropyl) Phosphate (TRIS). The Siterepresented a threat to public health, welfare, and the environment because of widespreadcontamination caused by poor waste management practices and direct discharge to the PineRiver. In 1982 U.S. EPA and the State of Michigan entered a Consent Judgment with Velsicolfor the Site. In the Consent Judgment Velsicol agreed to contain in place the 52 acre main plantsite where the former chemical plant was located. The parties to the Consent Judgmentconcluded at that time the most appropriate alternative for the Pine River sediments was to leavethe sediments in place, and the Consent Judgment released Velsicol from liability for clean up ofthe sediments that were contaminated at the time of entry of the Consent Judgment or becamecontaminated from migration or discharge from the main plant site prior to completion of theContainment System. This Record of Decision (ROD) concerns only OU2, Pine River sedimentand fish contamination.
The Site, including the Pine River, has been the subject of a number of investigationsconducted by the Michigan Department of Environmental Quality (MDEQ), U.S. EPA andVelsicol. The studies revealed Site soils contaminated with PBB, HBB, TRIS and othercontaminants; ground water contaminated with vinyl chloride, toluene, chlorobenzene, DDT andother contaminants; Pine River sediments contaminated with PBB, HBB and DDT; and elevatedlevels of PBB, DDT and other contaminants in fish. Pine River surface water did not containmeasurable levels of contaminants associated with the Site.
Upon completion of the Site characterization in the early 1980's, the State of Michigan,U.S. EPA, and Velsicol negotiated an agreement that included a remedy directed at stopping the
migration of PBB, HBB, DOT and other contaminants found at the Site into the environment.The remedy selected consisted mainly of ? 2 foot thick, low-permeability slurry wall around the52 acre facility and a 3 foot thick, low-permeability, clay cap. Under the Consent JudgmentVelsicol must maintain ground water levels within the slurry wall and cap ("ContainmentSystem"). Construction of the Containment System was completed by Velsicol in 1984. TheSite was proposed for inclusion on the National Priorities List (NPL) on December 30, 1982, andappeared on the final NPL on September 8, 1983.
The 52 acre main plant site is now covered with shallow-rooted grass, and, to restrictaccess, enclosed by a chain link fence. Velsicol i; currently operating and maintaining the Site inaccordance with an approved operation and maintenance plan requiring weekly inspections forsigns of deterioration, quarterly monitoring of gas vents, measurement of groundwater levelswithin the contained Site, and slurry wall permeability testing.
The Consent Judgment did not require Velsicol to remove the contaminated sedimentsfrom the Pine River/St. Louis Impoundment. A 1988 Preliminary Health Assessment preparedby the Michigan Department of Public Health (MDPH) and the Agency for Toxic Substances andDisease Registry (ATSDR) concluded the river poses a potential public health concern becasueof possible human exposure to contaminants via ingestion offish and direct contact with riversediments. The concern with potential fish consumption was reiterated in 1993 in anMDPH/STSDR Site Review and Update. Contamination of fish in the river was addressed byhealth advisories issued by the State of Michigan. A no consumption advisory for all species offish was initially published in the Michigan fishing guides in 1977, and is presently in effect.The no consumption advisory affects 33 river miles of the Pine River.
Water levels inside the Containment System (slurry wall and cap) remained below thelevel set by the 1982 Consent Judgment until February 1993. In 1993 Velsicol had to pump 1.25million gallons of water from the Containment System to stay below the established level. In late1994 Velsicol removed another 1.28 million gallons of ground water from the system to maintainthe level set in the Consent Judgment. Velsicol has continued to pump water from theContainment System approximately every 6 months to maintain the required water level, andVelsicol has been disposing of this water off-Site. Meanwhile the State collected fish samples inlate 1994 and noted that the average concentration of total DOT in skin-off filet carp samplesmore than doubled since the last collection in 1989. Average concentration of total DOT in 1989was 10.5 ppm, in 1994 tissue concentrations were 23.3 ppm. The State collected fish again in1995 and found an average total DDT concentration in skin-off filet carp samples of 16.1 ppm.The contaminant concentrations in fish tissue coupled with the water intake to the ContainmentSystem caused concern that the Containment System may have failed increasing the loading ofDDT into the Pine River.
Velsicol agreed to reassess the Containment System to ensure that it was not a source ofDDT into the Pine River. At the same time U.S. EPA and MDEQ (the Agencies) reassessedsediment contamination in the Pine River and decided to reconsider the no action decision madein 1982.
In 1996, Velsicol completed a comprehensive assessment of the Containment System.Velsicol's assessment of the clay cap included collection of samples from the upper portion ofthe cap and analysis for permeability, grain size, and Atterberg limits. Assessment of the
containment wall consisted of installation of inclinometers inside and outside the slurry wall atseven locations, installation of settlement plates at seven locations inside the slurry wall,
"collection of samples at nine locations for permeability analysis; installation of upper zonepiezometers on the inside and outside of the wall at five locations; water level measurements andfree product screening from all monitoring wells and piezometers; and a dye tracer study at thefive locations where the piezometers were installed. Velsicol published a report entitled FinalContainment System Assessment Report, Former Michigan Chemical Plant Site, St. Louis,Michigan, October I, 1997 detailing the Containment System assessment and results.
The Agencies believe the results of the Containment System Assessment document thatthe clay cap is leaking, probaMy because there is no frost protection layer on top of the cap. Noobvious problems were documented with the slurry wall. Velsicol concluded in their report ofthe findings that the Containment System is working as designed. On December 11,1997Velsicol submitted a work plan entitled Work Plan Post-Closure Cap Maintenance, FormerMichigan Chemical Plant Site, St. Louis, Michigan in which Velsicol states it will conductmaintenance of the clay cap during summer, 1998 by recoinpactmg areas of tlie clay cap.Velsicol decided to delay this work until U.S. EPA and MDEQ completed the sediment removalproject.
Simultaneously with the Containment System Assessment, the Agencies began areassessment of contamination in the Pine River/St. Louis Impoundment. During summer, 1996sediment cores were collected from 23 locations in the St. Louis Impoundment and analyzed forPBB, HBB and DDT. Surficial sediment samples were also collected from depositional areas inthe lower Pine River (below the St. Louis dam). During summer, 1997 the Agencies collectedanother round of sediment cores from 28 locations and analyzed them for DDT and total organiccarbon (TOC). MDEQ collected fish for analysis.
On June 8, 1998 U.S. EPA signed an Action Memorandum for a time-critical removalaction at the Site. The removal action consists of dredging/excavating sediments containing3,000 ppm total DDT or greater (the hot spot), treatment of the sediments with astabilizing/drying agent and disposal of the sediments off-Site. The removal action also includesbuilding the infrastructure necessary to complete the removal action such as roads, staging pad,and water treatment plant. Construction of the infrastructure was substantially complete byNovember, 1998. Sediment removal is expected to begin in spring, 1999.
Ill Community Participation
The streamlined Remedial Investigation/Feasibility Study (RI/FS) Report and Proposed Planfor the Velsicol Site were made available to the public in August, 1998, They can be found inthe Administrative Record file and the information repository maintained at the U.S. EPAdocket room in Region 5 and at the T.H. Cutler Memorial Library in St. Louis, Michigan.The notice of the availability of these two documents was published in the Alma Morning Sunnewspaper on September 9 and 13, 1998. A public comment period was held from September8, 1998 to October 8, 1998. A request for an extension to the public comment period was notreceived by U.S. EPA. A public meeting was held on September 16, 1998 to present theProposed Plan to a broader community audience than those that had already been involved at
the Site. At this meeting, representatives from U.S. EPA and the MDEQ answered questionsabout problems at the Site and the remedial alternatives. U.S. EPA's responses to thecomments received during this comment period are included in the Responsweness Summary,which is part of this ROD.
IV. Scope and Role of Operable Unit
As with many Superfund sites, the problems at the Velsicol Chemical Site are complex. As aresult, U.S. EPA has organized the work into two operable units (OUs):
• Operable Unit 1: Main plant site soils and gioundwaterOperable Unit 2: Contaminated sediments and fish in the Pine River and St. Louis
Impoundment
U.S. EPA, the State of Michigan and Velsicol Chemical Co, already agreed to aremedy for Operable Unit 1 in a 1982 Consent Judgment. This remedy, on-Site containmentof contaminated soils and groundwater was implemented by Velsicol in 1984.
The second operable unit, the subject of this ROD, addresses the contaminatedsediments in the St. Louis Impoundment of the Pine River. Ingestion of fish from the PineRiver poses a current and potential risk to human health because U.S. EPA's acceptable riskrange is exceeded. This second operable unit presents the final response action for this Siteand addresses a principal threat at the Site through the removal, treatment and disposal ofcontaminated sediments.
V. Characteristics of Pine River
The Pine River is part of the Saginaw River/Saginaw Bay drainage basin with a totaldrainage area of 312 square miles. The Pine River flows northeast toward Midland for 20.5miles where it discharges into the Chippewa River and then into the Tittabawassee River. TheTittabawassee then flows southeast toward Saginaw where it discharges into the Saginaw River.The Saginaw River then flows north where it empties into Saginaw Bay in Lake Huron(Figure 2-2).
The Pine River is impounded in both Alma and St. Louis. The Alma dam is locatedapproximately 4-5 miles upstream from the Site. The St. Louis Impoundment is adjacent to theSite arid immediately downstream of the Site is the St. Louis dam.
The portion of the Pine River and St. Louis Impoundment subject to consideration in thisdocument includes the Pine River from approximately the M-46 bridge to the Mill street bridge.The total area of this stretch is approximately 25 acres. The St. Louis Impoundment is identifiedin Figure 2-1 as the middle and lower basins and also identifies the areas described as the MainPlant Site, the Pine River, the M-46 bridge, the Mill Street bridge and the St. Louis dam.
The Pine River is a navigable waterway. Current uses of the Pine River and St. LouisImpoundment are impaired due to the sediment contamination. While sportfishing is not strictlyprohibited, anglers are limited to catch and release fishing by the no consumption fish advisory.
10
The advisory, however, is not easily enforceable. Swimming and boating are consideredundesirable due to the contamination. Generation uf electricity is currently the only acceptableuse of the river and impoundment. The impoundment provides hydraulic head for powergeneration. The City of St. Louis estimates that from July 1996 to June 1997 the St. Louis damgenerated 1.35 million kilowatt-hours of electricity, 4% of the total kilowatt-hours sold to theCity of St. Louis during that same time frame.
The closest gauging station to the Site is located on the right bank of the Pine River, 270feet downstream from the Superior Street Bridge in Alma. The location is 0.6 mile downstreamfrom the Alma dam and 5.2 miles upstream from the St. Louis dam. (USGS End of YearReport. 1993V
Within the St. Louis Impoundment, the water depth to sediment is generally between 7and 10 feet. The maximum observed depth during June 1993 sampling was 12 fee! near the inletto the hydroelectric plant.
The City of St. Louis participates in the national Flood Insurance Program. A FloodInsurance Study (FIS) was prepared for the Pine River in the City of St. Loui? by the FederalEmergency Management Agency (FEMA) in 1989. The FIS indicates that flooding in the City ofSt. Louis is primarily caused by overflow of the Pine River, however, the potential for flooddamage is not great because of the steep banks and flood elevation regulation provided by the St,Louis dam. The FIS presents the methodology and results of hydrologic and hydraulic analysesperformed for the Pine River to determine water-surface elevations corresponding to the 100-yeardischarge both upstream and downstream of the dam. Upstream of the dam, where thecontaminated sediments are located the flood profile is based on a starting water surfaceelevation at the dam that was derived from a series of 38 annual maximum pool elevations asprovided by the dam operator. The U.S. Federal Highway Administration's WSPRO model wasused to compute water surface elevations on the Pine River for the St. Louis FIS.
Sediment Data Summary
Sediment data was collected from the Pine River/St. Louis Impoundment in 1980, '81,'96 and '97. This summary presents only the 1997 data. Sediment data from 1996 was used toevaluate the lower basin because an equipment failure precluded the collection of sediment coresfrom that area in 1997. Body burden of contamination in fish has been monitored by MDEQsince 1983. Fish collection and analysis were completed in 1983, '85, '94, '95 and '97. Adetailed summar, , fall historic fish surveys is presented in this summary because ingestion ofcontaminated fish is the main exposure pathway. The risk assessment in the RI/FS Reportidentifies PBB, total DOT, and HBB as chemicals of concern for the Site.
In July of 1997, the U.S. EPA and MDEQ conducted sediment sampling in the Pine Riverand the Impoundment area. The survey was intended to supplement the May 1996 survey andprovide additional information regarding the nature and extent of the DDT contamination in thePine River. A total of 28 cores and 3 grabs were collected and partitioned into 0 to 6, 6 to 30, 30to 54 and 54 inches to refusal depth intervals. A total of 77 samples were analyzed for total DDTcontamination and total organic carbon (TOC). All sample locations are identified onFigure 2-10. On the last day of the survey the sampling equipment malfunctioned; therefore
11
only grab samples (surficial) were obtained instead of core samples in the lower basin at Stations27, 28, 29, 30, and 31.
The horizontal extent of total DOT in sediments of the Pine River and St. Louisimpoundment was estimated based on the 1996 and 1997 data as explained above. Themaximum concentration of total DOT ai each sample location was evaluated. Interpolation ofthe maximum concentration was completed to delineate the surface area of total DDTconcentrations exceeding 1, 10, 100, 500, and 1,000 ppm. The surface areas and depth was usedto estimate the volume of sediment associated with each concentration (i.e., 1, 10, 100, 500 and1,000 ppm). The areas and volumes for each concentration are presented on Figures 2-3 through2-6 and in Table 2.2-3.
Table 2.2-3: Concentration, Area, Volume and Mass of DDT in St. Louis ImpoundmentConcentration (ppm)
1
10
100
500
1000
Area (acres)
65
31
20
12
6
Volume (cy)
516,650
260,330
169,000
104,200
48,000
Mass (Ib)
540,000
533,000
529,000
519,500
490,700
As shown in Table 2.2.1.4-1, sediment total DDT concentrations ranged from 1.3 to32,600 ppm. Of 77 samples, none of the samples indicated total DDT concentrations less than 1ppm, 44 were between 1 and 10 ppm. 14 were between 10 and 50 ppm, 3 were between 50 and100 ppm, 14 were between 100 and 1000 ppm, and 7 samples were greater than 1000 ppm totalDDT.
Table 2.2.1.4-1: Total DDT Concentration per Depth Interval (1997 Data)Depth Interval (in.)
0 - 6
6-30
30-54
54-112
Max. Cone, (ppm)
229.
32,600*
32,600*
822
Avg. Cone, (ppm)
28.4
1485
1639
130
Min. Cone, (ppm)
1.3
1.8
2.1
2.4* The maximum concentration was found in a sample collected from 6 to 42 inches.
The average TOC content of the impoundment surface sediment was 3.1 percent. Theaverage TOC content of subsurface samples was 3.5 percent.
Based on the data obtained from the sediment surveys several observations can be made.First, the location of the maximum total DDT concentrations were consistently found in the
12
sediment depth interval ranging from 6 to 42 inches. This area includes the Impoundment fromthe Site plant jetty to the Mill Street bridge in the middle basin.
Results from all sediment surveys indicate that the Isvels of total DOT in the Pine Riverand the St. Louis Impoundment are extremely high. Analyses from the 1980, 1981, 1996, and1997 data show that the concentration levels, as a whole, have not decreased over time.
Five sediment grab samples were collected by MDEQ in the Pine River between the St.Louis dam and the confluence with the Chippewa River. Significant amounts of total DOT werenot found in sediments below the St. Louis dam. The results are summarized in Table 2.2.13-2.
Table 2.2.1.3-2; Total DOT Concentrations in Sediment below St. Louis DamLocation
McGregor Road
downstream of Bagley Road
Magrudder Road
9 Mile Road
Meridian Road
Total DOT (ppm)
0.566
0.117
0.258
0.143
none detected
Fish Data Summary
Fish tissue samples were collected from the St. Louis Impoundment and below the St.Louis dam by the MDEQ in 1983, '85,489, '94, '95 and '97. Generally, fish tissue samples wereanalyzed for DDT, DDE, ODD and PBB. This summary of fish data focuses on trends in skin-off filet (Fs) carp samples since this was the only sample type consistently collected. Table2.2.2-1 summarizes the maximum, average and minimum concentration of total DDT in skin-offfilet samples of carp. Table 2.2.2-2 illustrates specific sampling factors that may explainvariability in the data and highlights data trends.
Table 2.2.2-1: Skin-off Filet Carp SamplesCollection
Date
1983*
1985
1989
1994
1995
1997
Species
Carp
Carp
Carp
Carp
Carp
Carp
Type
Fs
Fs
Fs
Fs
Fs
Fs
Max. Cone.(ppm)
0.10
18.66
39.80
4730
43.30
8992
Avg. Cone, (ppm)
0.06
9.66
10.50
23.30
16.10
34.57**26.82**
Min. Cone.(ppm)
0.03
5.27
0.06
1.58
0.50
2.47
Fish collected in 1983 were not analyzed for DDD and DDE
13
** 34.57 ppm total DOT is the average concentration for fish collected in the St. LouisImpoundment, 26.82 ppm total DOT is the average concentration for fish collected belowthe St. Louis dam.
Table 2.2.2-2: Carp Data Specifics
op
dde
avg
PP
dde
avg
op
ddd
avg
PP
ddd
avg
op
ddt
avg
PP
ddt
avg
% fat
avg %
fat
min %
fat wt
avg wt mm
wt
location
1983
(8)
.08 3674 2403 680 below dam
1989
(10)
3.1 6.1 .29 6.65 2.08 1810 1149 820 Impound
1995
(10)
4.9 10. .17 4.1 2.09 0.6 4360 1789 920 Impound
1997
8.6 6.2 15. .97 10.4 3.15 1.10
llltli
5595 3099 1900 Impound
(8)
Each of the breakdown products that make up total DDT (DDD, DDE and DDT) havetwo isomers, ortho para ("op") and para para ("pp").
In 1985 all eight carp samples obtained below the dam exceeded the MichiganDepartment of Community Health (MDCH) and Federal Food and Drug Administration (FDA)tolerance level of 5 ppm. Eight of the 10 carp samples collected in 1994 and all 12 of the carpsamples collected in 1997 (below the dam) exceeded MDCH and FDA tolerance level of 5 ppmfor total DDT in fish.
In 1989 four often carp samples collected in the Impoundment exceeded the MDPH andFDA tolerance level of 5 ppm. Six often carp samples collected in 1995 and six of eight of the
14
carp samples collected in 1997 (in the St. Louis Impoundment) exceeded the 5 ppm tolerancelevel for total DOT in fish.
It is difficult to draw conclusions from the data due to variabilities (weight, age, % fat,number of samples collected, etc.), however, what is clear is that carp in the St. LouisImpoundment and below the St. Louis dam are bioaccumulating high levels of total DOT andthere are no indications, from the 14 years of data, of a downward trend. If any conclusion couldbe drawn it would be that contaminants in sediments continue to be increasingly bioavailable tofish. Fish in the Pine River are highly contaminated and will continue to be if no remedialaction is taken at this Site.
Surface Water Data Summary
Fourteen water samples were collected in June, 1980 from the Pine River between theCheeseman Road bridge and the St. Louis Municipal wastewater treatment plant outfall todocument the contamination of the Pine River. These samples were collected in one gallon glasscontainers using a battery operated vacuum pump with teflon tubing. The samples were testedfor PBB, HBB, DOT and its analogs, and subjected to a limited organic scan.
The Pine River water samples did not contain measurable concentrations of HBB, PBB,or DOT analogs. Elutriate testing under laboratory conditions showed that these contaminantsare bound tightly with the sediment and do not readily desorb and solubilize from the sedimentsto the water. Only a small amount of p,p-DDT was detected in the water from the static phase ofthe testing. The amount desorbed was less than 0.01% of the amount present in the sediment.
Two water samples, RWS-6-01 and RWS-6-02, did contain 3.5 Mg/1 and 2.7 ^g/1methoxychlor, respectively. The origin of this compound in the samples is not known.
In October of 1992 water sampling was again conducted. The City of St. Louis had awater quality study completed in the area surrounding the St. Louis dam. Six samples werecollected by Ayers, Lewis, Morris, & May, Inc. Three of these samples were taken above thedam, and three samples were taken below the dam. The samples were tested and reported as twocomposites, above and below the dam.
The laboratory results showed levels of total DOT, other hydrocarbons and pesticides,and BTEX for the two composites to be below the method detection level. While most heavymetals were at or below the analytical detection limit, the City sampling event showedinsignificant treaces of some heavy metals.
VI. Current and Potential Future Site Uses
The portion of the Velsicol Site currently under investigation is the St. LouisImpoundment. Due to the high levels of contamination in the sediment of the St. LouisImpoundment and in fish in the Pine River the only current acceptable use of the river isgeneration of electricity. However, U.S. EPA is aware that some local residents fish despitethe no consumption fish advisory and there is a migrant farm worker population that comes toSt. Louis every summer to live in camps and work on the surrounding farms. U.S. EPA's riskassessment assumes that this population will fish the Pine River to supplement their diet, and
15
that fish eating birds are fishing in the Pine River.If the clean-up goal is attained, uss of the river and impoundment should be unrestricted
in the future. Eventually, fish are expected to be clean enough for humans and birds to eat.All types of recreational activities would be available, swimming, wading and canoeing. Theimpoundment will continue to provide hydraulic head for the generation of electricity.
VII. Summary of Site Risks
This section presents a summary of the key findings of the baseline risk assessment forOU2. The risk assessment in the RI/FS Report identifies PBB, total DOT, and HBB aschemicals of concern for the Site. The risk assessment determined that the main exposurepathway is through ingestion of contaminated fish. Chemicals detected in fish in 1997 above thedetection limit were mercury, DOT and its metabolites, chlordane congeners, PBB,hexachlorobenzene and octachlorostyrene. The DOT concentration in most of the fi«h collectedexceeded the Michigan Department of Community Health (MDCH) Level of Concern of 5 ppmtotal DOT. The mercury and chlordane concentrations did not exceed the MDCH Levels ofConcern. MDCH has no official Level of Concern for the other chemicals detected.
Risk associated with dermal contact considered absorption of total DDT, PBB, and HBB.
Human Health Risks
The baseline risk assessment estimates what risks the Site poses if no action is taken.The baseline risk assessment provides the basis for taking action and identifies thecontaminants and exposure pathways that need to be addressed by the remedial action.
DDT and its breakdown products DDD and DDE, PBB and HBB are the chemicals ofconcern for this OU. Due to their hydrophobic and lipophilic characteristics, the contaminantsare partitioned in sediments and bioaccumulated in fish tissue. DDT, DDD and DDE werefound at the highest concentrations in fish tissue and sediments and therefore are the basis forthe human health and ecological risk assessments. Because DDT, DDD and DDE areextremely bioaccumulative, the chief exposure threat is through the food chain, fromconsuming contaminaed fish from the Pine River. People may also be exposed to contaminantsduring wading (or other recreational activities) due to direct contact of their skin withsediment. While this may not occur frequently, it is thought to occur to some degree given thevery close proximity of many homes to the river and lack of any appreciable barriersseparating backyards from the River (including vegetation, presence of any banks, dock faces,etc.). Other exposure routes such as inhalation or surface water ingestion are thought to beminimal and insignificant compared to ingestion of fish and direct contact with sediments,because of the low level volatility of DDT and it's strong partitioning to sediment. Theseexposure routes were not assessed further in the risk assessment.
Exposure scenarios are defined by the exposure assumptions used, Table 1 lists allexposure assumptions used in all scenarios. In this analysis, the critical assumptions that helpdefine scenarios were:
* ingestion rate (how many grams a day of fish people consume),
16
* fraction ingested (how much fish consumed comes from Pine River)- • type of fish consumed (sport fish vs. bottom feeding fish)A study done by West et al. on Michigan anglers provides useful data on ingestion rates for thevarious scenarios. For the lower bound estimate, the SCf percentile ingestion rate from thisstudy was used (a daily average of 20 grams per day). For the reasonable maximum exposure(RME), the 95th percentile ingestion rate was used (a daily average of 75 grams). Given thequality and specificity of this study for use at this Site, there is a good deal of confidence thatthese exposures are reasonable and do occur for some portion of the population. For the RMEscenario, a study of a three day maximum fish consumption was used instead of West data.This again is in line with a worst case scenario. For fraction ingested from Pine River, studyassumptions were made that reflect the different scenarios. For the lower bound or sportfishing scenario, 50% was used, meaning of the total amount of fish consumed, 25% comesfrom the Pine River. In the RME U.S. EPA assumed that 50% of fish consumed comes fromthe Pine River, and for the subsistence scenario U.S. EPA assumed that all fish comes fromthe Pine River.
|TABLE 1. Exposure Assumptions
Body weight (kg)
I Ingestion rate (kg/day)iii! Fraction ingestion (%)! Absorption (%)i Exp. frequency (days/year)| Exposure duration (years)
SPORT RME SUBSIS.70 i
25550:
0.075
The recent sediment data collected in July 1997 and ash data collected in October 1997 wereassessed to estimate current risks of consumption of contaminated fish and dermal risks. Bothcancer effects and noncancer effects of DOT were assessed. The noncancer effect beingassessed here is liver lesions.
Toxicity
DOT and its breakdown products (ODD and DDE) are classified as probable human
17
carcinogens based on liver tumors, lung tumors and thyroid tumors in rodents. DDT, ODD andDDE are all structurally similar. Both hepatocellular adenomas and carcinomas were observed insix mouse liver tumor studies upon treatment with DDT. The cancer slope factor for DDT, asobtained from IRIS (1997) is 0.34 (mg/kg-day)'1. Treatment of mice with DDD resulted in astatistically significant increase in incidence of lung tumors in both sexes of mice whencompared to controls. The cancer slope factor for DDD is 0.24 (mg/kg-day)"1. DDEadministered to mice resulted in a dose-dependent and statistically significant increase inincidence of hepatocellular carcinomas in both males and females in comparison to controls.The cancer slope factor for DDE is u.34 (mg/kg-day)'1.
A slope factor for assessing cancer risks assumes th^t cancer risk is probabilistic and anydegree of exposure leads to some degree of risk. A slope factor relates estimated exposures toincremental lifetime cancer risks, and therefore the result is a probability of cancer over thebackground levels in the population. Therefore a risk result of 7 E-4 is equivalent to saying thereis an increase cancer risk at a rate of 7 in 10,000 people. This assessment combines the data forDD f and its breakdown products", DDD and DDE, and applies the cancer slope factor of 0.34.
DDT and its breakdown product (DDD and DDE) have also been reported to exert non-cancer effects. IRIS specifically calculates a Reference Dose (RfD) for DDT based upon livertoxicity in rats. The RfD was based upon a study performed in adult male rats and is 0.0005mg/kg-day. An RfD is intended to indicate a safe level exposure, meaning that exposure at theRfD level is likely to be without an appreciable risk of deleterious effects. To assess non-cancerrisks, a hazard index of the estimated exposure over the RfD is calculated. Because the RfDrepresents a safe level, the hazard index should be one or less than one. The higher the hazardindex the higher the likelihood of adverse effects.
In addition, DDT and its breakdown products (DDD and DDE) are undergoing increasingscrutiny for their role as endocrine disrupting compounds. As endocrine disrupters, DDT has thepotential to negatively impact the developing fetus, increase vulnerability to certain cancers, andpossibly decrease fertility. While conclusive studies have not been done in order to develop anRfD based on these endpoints, the current evidence suggest some endocrine disrupting potentialfor DDT and its breakdown products. Due to the tremendous uncertainty around this endpoint,this risk assessment applies the RfD for DDT (calculated based upon liver toxicity in adult malerats) to the breakdown product of DDT (DDD and DDE) for which no RfD has been calculated.
Polybrominated biphenyls (PBB) are classified as probable human carcinogens based onliver tumors in rats. Both carcinomas as well as neoplastic nodules were observed in rats treatedwith PBB. The cancer slope factor for PBB, as obtained from HEAST (1995), is8.9 (mg/kg-day)1.
PBB has also been reported to exert non-cancer effects. HEAST reports an RfD for PBBbased upon liver toxicity in rats as 7 x 10"6. In addition, PBB is undergoing increasing scrutinyfor its role as an endocrine disrupting compound. As an endocrine disrupter, PBB has thepotential to negatively impact the thyroid, the developing fetus, increase vulnerability to certaincancers, and possibly decrease fertility.
Hexabromobenzene (HBB) does not have a carcinogenicity assessment entered into eitherIRIS or HEAST. An RfD has been calculated for HBB based upon induced serum carboxylesterase activity and increased liver-to-body weigM ratio. The RfD for HBB is 0.002 mg/kg-day.
18
The confidence in the RiD is low because the critical study was of short duration, only one sex(male) was exposed, and few definitive parameters were examined.
The hazard indices from all three chemicals can be appropriately added because of theshared target organ of the liver.
Risk CharacterizationFish consumption
Risks were estimated by using the 1997 fish data, both the smallmouth bass and carp datawere used, along with the exposure assumptions for each scenarios described in Table 1. Asummary of the risks for each species for each exposure scenario is shown in Table 2.
TABLE 2. Current Risks from Fish Consumption
Smallmouth bassCarp
Cancer sport3.6 E-051 E-04
RME3 E-041E-03
subs is.4E-03
1.6E-02
Hazard sport0.210.63
RME1.76
subs is.22.695
Table 2 clearly shows risks are within or exceeding the target risk range of E-4 - E-6for all exposures and all species. Even consumption of bass, which is frequently much lesscontaminated than other fish, is of significant concern. The RME estimate for both speciesshows that for any combination of fish caught, risk increases and even recreational fishing withonly 50% coming from this Site is not without risk.
Dermal Risks
Risk of cancer via dermal contact with the sediments was estimated for a typicalexposure using concentrations in the top 6 inches of sediment and a reasonable maximumexposure, using the maximum at subsurface levels of sediment. Table 3 shows current risksfrom dermal exposures using 1997 data. Table 3 shows the current risk from dermal exposureusing 1997 data.
TABLE 3. DERMAL CANCER RISKS USING 1997 DATA
Typical (0-6 in. sediment)
2.0E-05
RME (maximum at subsurface)
2.0E-02
19
Ecological Risks
This section presents a summary of the key findings of the ecological risk assessment forOU2 as presented in the RI/FS Report. The results of the ecological risk assessment show thatfish-eating birds, as represented by great blue heron, that consume fish from the St. LouisImpoundment are at risk for reproductive impairment related to eggshell thinning and otheradverse effects caused by DDE. The calculated dose of DDE to heron (2.04 mg DDE/k&w-d)is 3 times greater than the DDE lowest observed adverse effect level (LOAEL) of 0.7 mgDDE/kgBW-d, and 30 tunes greater than the no observed adverse effect (NOAEL) of 0.07 mgDDE/kgBW-d (Table 2.4-6). Adverse reproductive effects would therefore be expected inheron that obtain one-third or more of their diet from the St. Louis Impoundment. Adverseeffects are unlikely to occur in fish-eating birds that obtain no more than 3% of their diet fromthe St. Louis Impoundment.
The preliminary remedial goal (PRO) for sediments is 1 ppm DDT to be fullyprotective of reproduction in fish-eating birds. This sediment concentration should result in0.5 mg DDE/kg whole fish, the dietary concentration that corresponds to the NOAEL.Adverse reproductive effects may be expected when mean sediment levels exceed 8 to 12 ppmDDT, which should result in about 5 mg DDE/kg whole fish (LOAEL). These levels arebased solely on the adverse effects associated with DDE, and do not consider the potentialadditive effects associated with DDD or unmetabolized DDT.
Concentrations of DDT in St. Louis Impoundment sediments exceed sediment screeningvalues for potential adverse effects on benthic invertebrates. Two-thirds of the surfacesediment samples also exceed the severe effect levels (SEL) that indicate a potential for adverseeffects on the majority of benthic organisms. Forty percent of the surface sediment samplesexceed the threshold for effects on benthic populations as determined at another Site. Threesamples from the middle basin exceed the median lethal concentration (LQo) determined insediment toxicity tests at another Site.
Although these comparisons do not prove that the contamination in the St. LouisImpoundment are adversely affecting benthic invertebrates (because Site-specific studies havenot been performed), they indicate that adverse effects are likely and support the conclusion forfish-eating birds that the Site represents a significant ecological risk.
Basi" for Action
Actual or threatened releases of hazardous substances from this Site, if not addressed byimplementing the response action selected in this ROD, may present a current or potentialthreat to public health, welfare, or the environment.
VIII. Remediation Objectives
This section identifies the Site-specific Remedial Action Objectives (RAOs). TheseRAOs pertain to "general Site cleanup" or are intended to fulfill potential federal and stateApplicable or Relevant and Appropriate Requirements (ARARs) and "to be considered"
20
criteria (TBCs). The RAOs proposed for this Site, where DDT and its breakdown productsare the primary constituents of concern, are as follows:
• Reduce DDT concentrations in fish and sediments in the St. Louis Impoundmentto levels that would not present an unacceptable human- health or ecological riskand would allow eventual elimination of existing fish consumption advisories;
• Prevent direct human contact with contaminated sediments;• Prevent significant down river migration of contaminated sediments;• Achieve compliance consistent with federal and state ARARs for the Site; and• Comply with risk-based objectives defined by the risk assessment.
IX. Description of Alternatives
U.S. EPA Region 5 is recommending waste removal alternative 4 and disposalalternatives 5 and 6. Alternative 4 consists of hydraulic modification of the Pine River,excavation of sediments above 5 ppm total DDT (approximately 260,000 cu. yds.).Alternative 5 consists of disposal of contaminated sediment in a RCRA subtitle D landfill andAlternative 6 consists of disposal of contaminated sediment in a RCRA subtitle C landfill.Each of the alternatives considered are described below. The long-term operations andmaintenance costs were calculated for a 30 year period at an 8% discount rate.
Alternative 1: No Action - This alternative involves taking no additional action at the Site, butincludes fish tissue monitoring and fish advisories that are currently in place. This alternativeserves as a baseline against which other alternatives are evaluated.
Waste Removal Alternatives
Alternative 2A: Hydraulic Dredging, Dewatering and Water Treatment - Alternative 2Aconsiders hydraulic dredging of sediments with DDT concentrations in excess of the identifiedaction level (approx. 260,000 cu. yds.), dewatering of sediments, addition of a stabilizingagent, water treatment and discharge to the Pine River/St. Louis Impoundment. The watertreatment will be designed to meet surface water discharge requirements and will probablyconsist of clarification, sand filtration, carbon filters and 1 micron bag filters. Duringdredging operations monitoring will be conducted to ensure protection of workers and thecommunity. Monitoring of resuspension will be done directly downstream of the dredgingoperations. Air monitoring will be conducted. After completion of the dredging projectsediment samples will be collected to ensure the clean-up standard has been met. The State ofMichigan will continue to monitor fish tissue levels.
Alternative 3 A: Mechanical Dredging, Dewatering and Water Treatment - Alternative 3Aconsiders gasketed clamshell dredging of sediments with DDT concentrations in excess of theidentified action level (approx. 260,000 cu. yds.), dewatering of sediments, addition of astabilizing agent, water treatment and discharge to the Pinr River/St. Louis Impoundment.
21
The water treatment will be designed to meet surface water discharge requirements and willprobably consist of clarification, sand filtration, carbon filters and 1 micron bag filters.During dredging operations monitoring will be conducted to ensure protection of workers andthe community. Monitoring of resuspension will be done directly downstream of the dredgingoperations. Air monitoring will be conducted. After completion of the dredging projectsediment samples will be collected to ensure the clean-up standard has been met. The State ofMichigan will continue to monitor fish tissue levels.
Alternative 4: Hydraulic Modification of the Pine River, Excavation of Sediments, DeWiueringand Water Treatment. Alternative 4 is similar to Alternative 3A except that temporarycofferdams will be placed in the St. Louis Impoundment, water will be pumped from thecofferdam, the sediments will be excavated instead of dredged, a stabilizing ageni will beadded and water will be treated prior to discharge to the Pine River. The water treatment willbe designed to meet surface water discharge requirements and will probably consist ofclarification, sand filtration, carbon filters and 1 micron bag filters. During excavationoperations monitoring will be conducted to ensure protection of workers and the community.Monitoring of resuspension will be done directly downstream of the excavation operations.Air monitoring will be conducted. After completion of the dredging project sediment sampleswill be collected to ensure the clean-up standard has been met. The State of Michigan willcontinue to monitor fish tissue levels.
Contaminated Sediment Disposal Options
Alternative 5: This alternative considers disposal of sediment in a commercially operatedRCRA subtitle D landfill located in the State of Michigan.
Alternative 6. This alternative considers disposal of sediment in a commercially operatedRCRA subtitle C landfill.
Containment Option
Alternative 7: This alternative considers capping all the contaminated sediments in place.Alternative 7 considers the placement of a sand cap with a stone armoring system consisting ofa 20-inch coarse-grained sand cap and 5- to 7.5-inch diameter stone armor layer. Monitoringwould occur every 2-3 years and cap replenishment would occur every 5 years.
22
Alternative
1 : No Action
2A. HydraulicDredging
3A: MechanicalDredging
4: Temporary Damsand Excavation
5: RCRA Subtitle DLandfill
6: RCRA Subtitle CLandfill
7: Capping
CapitalCosts
$0
S22.4M
S20.7M
$16.9M
$3.2M
S17.4M
$7.5M
O&M Costs/Year
$16,000
$0
$0
$0
$0
$0
$30,100
30 yr.PresentWorth Costs
$180,000
NotApplicable
NotApplicable
NotApplicable
NotApplicable
NotApplicable
$7.84M
Total Cost
$ 180,000
$22.4M
$20.7M
$16.9M
$3.2M
$17.4M
$7.8M
X. Summary of Comparative Analysis of Alternatives
Alternative I, the No Action alternative, would not provide adequate protection ofhuman health or the environment. The risk assessment documents that unacceptable riskwould occur to humans and fish-eating birds if fish are ingested from the Pine River. Theserisks cannot be adequately addressed through fish advisories, especially for fish-eating birds.Therefore, the No Action alternative acts as a baseline to compare Alternatives 2A, 3A, 4 and7 against. Alternative 5 and 6 cannot stand alone, and must be paired with one of the sedimentremoval options.
Alternatives 2A, 3A and 4 are very similar; they all entail the removal of contaminatedsediments above 5 ppm total DDT. The only difference between these alternatives is themethod for removal of the sediments. Alternative 2A considers hydraulic dredging,Alternative 3A mechanical dredging, and Alternative 4 considers excavation. All of theseremoval methods are implementable at this Site. Alternative 4 is preferable over alternatives2A and 3A because excavation is the most efficient way to remove sediments and will producethe least amount of water that will require treatment. However, U.S. EPA recognizes thatinstallation of temporary coffer dams may not be implementable in all locations in the St. LouisImpoundment and therefore some of the sediment removal may need to be completed usingmechanical or hydraulic dredging.
Alternatives 5 and 6, disposal in a Subtitle D and C landfills respectively, are bothconsidered to be favorable. Subtitle C landfill will be used to dispose of the highly
23
contaminated sediments and a Subtitle D landfill will be considered for disposal of less highlycontaminated sediments.
Capping the contaminated sediments in place, Alternative 7, is least favored due to thesignificant uncertainty with the effectiveness of this option at protecting human health and theenvironment. There are many unknowns with how effective a cap will be in reducingbioavailability of contaminants to fish. Physical disturbances, diffusion and advection all cancompromise the ability of a cap to perform adequately. In addition, capping is not a proventechnology in shallow water. Nationally caps have only been installed at deep water sites, andnone of these sites have post-monitoring data to .show that they have been effective at reducingcontaminant levels in fish. Dredging on the other hand has been shown through post-monitoringdata at the Waukegan Harbor site in Illinois to have reduced contaminant levels in fish such thata fish advisory was removed. At the Black River site in Ohio dredging resulted in reducednumbers of tumors in fish. Deposition of cleaner material and biodegradability are not effectiveat reducing contaminant levels in fish as evidenced by the fish data which clearly shows thatalthough contaminant levels in the top 6 inches of sediments is declining, contamination in fishtissues is not. If Alternative 7 was selected and implemented and then found not to be effectiveat reducing contaminant levels in fish tissue, U.S. EPA would have to reconsider the remedyand potentially select one of the removal alternatives instead. This would significantly increasethe costs overall, since not only would U.S. EPA have to remove the cap, but also thecontaminated sediments.
XI. Selected Remedy
U.S. EPA Region 5 is selecting Alternatives 4, 5 and 6: Hydraulic Modification of thePine River, Excavation of Sediments, Dewatering and Water Treatment and disposal ofcontaminated sediments in either a RCRA Subtitle D or C landfill. Alternatives 4, 5 and 6represent the best balance of the nine criteria. U.S. EPA recognizes that installation oftemporary coffer dams may not be implementable in all location in the St. LouisImpoundment and therefore some of the sediment removal may need to be completed usingmechanical or hydraulic dredging (Alternatives 2A or 3A).
Alternative 4 contemplates the use of temporary cofferdams and excavation ofsediments. The temporary coffer dams will be placed in the St. Louis Impoundment, waterwill be pumped from the cofferdam, the sediments will be excavated, a stabilizing/drying agentwill be added and wastewater will be treated prior to discharge to the Pine River. The watertreatment will be designed to meet surface water discharge requirements and will probablyconsist of clarification, sand filtration, carbon filters and 1 micron bag filters. Duringexcavation operations monitoring will be conducted to ensure protection of workers and thecommunity. Monitoring of resuspension will be done directly downstream of the excavationoperations. Air monitoring will be conducted. After completion of the dredging projectsediment samples will be collected to ensure the clean-up standard has been met. The State ofMichigan will continue to monitor fish tissue levels.
For a detailed cost estimate of the selected remedy, see Appendix B of this ROD.
24
Expected Outcomes of the Selected Remedy
U.S. EPA utilized a volume break point analysis and risk reduction analysis to determinea protective and cost effective cleanup goal for total DOT in the Pine River. Table 2.5-1 shows arange of cleanup goals (1, 5, 10 and 100 ppm total DOT) and compares the area, volume ofsediment required to be removed, and total DDT mass that would be removed if the goal weremet. The average dredging depth is assumed to be 5 feet.
T?ble 2.5-1 Volume Break Point AnalysisCone, (ppm)
1
5
10
100
Area (acres)
65
32
31
20
Volume (cy)
516,650
262,380
260,330
169,900
Mass (Ib)
538,730
534,305
533,000
529,000
Table 2.5-2 shows the concentration of total DDT that would remain in fish tissue andreduction in risk if a cleanup goal of 1 or 5 ppm total DDT were met.
Table 2.5-2 Post Remedial Risk
Cleanup Goal
1 ppm
5 ppm
Concentration in Fish
Smallmouth Bass
0.5 ppm
0.8 ppm
Concentration in Fish
Carp
1 .0 ppm
1 .7 ppm
RME Risk
Smallmouth Bass
1 . 1 E-05
1.9E-05
RME Risk
Carp
2.3E-05
4. t E-05
Removing sediment with DDT at or above 5 ppm, will reduce levels of DDT in fishtissue by over 95% from current levels of 12.5 to 0.8 ppm in Bass and 42.5 to 1.7 ppm in Carp.A 1 opm cleanup doubles the total volume of sediment to be removed while providing only anadditional 0.3 ppm reduction in fish tissue levels (12.2 to 0.5 ppm). Therefore, removingsediment in the Pine River contaminated with 5 ppm DDT and higher, will obtain the maximumreduction of risk to human health and the environment that is practicably achievable at the Site.To attain a 5 ppm level in sediments approximately 260,000 cubic yards of sediments would beremoved from the St. Louis Impoundment. The removal action together with the remedial actionwill accomplish removal of sediment with total DDT above 5 ppm.
25
XII. Statutory Determinations
This section discusses how the Selected Remedy (Alternatives 4, 5 and 6) satisfies thestatutory requirements of CERCLA section 121 (as reflected in the NCP requirements at 40CFR Section 300.430(f)(5)(ii) and explains the five-year review requirements for the SelectedRemedy.
Overall Protection of Human Health and the EnvironmentAlternative 4 considers removing all sediments at or exceeding 5 ppm total DDT from
the St. Louis Impoundment using temporary cofferdams and excavation. Based on the humanhealth risk assessment and ecological risk assessment, lemoval of sediments at or above 5 ppmwill provide protection to human health and the environment because risk associated with theexposure pathways will be substantially reduced. Removal of contaminated sediments areexpected to produce a significant decrease in the levels of total DDT in fish tissue, resulting insignificantly reducing the possibility of human health effects from consumption ofcontaminated fish and reproductive effects to fish-eating birds. The second exposure pathway,direct contact with contaminated sediments would also be eliminated by removing sediments ator above 5 ppm. This alternative will meet all the remedial action objectives (RAOs) and willnot pose unacceptable short-term risks.
RCRA subtitle D landfills are constructed for the purpose of disposal of municipal solidwaste. Construction requirements for bottom liners, leachate collection and monitoring aregenerally not as stringent as for landfills designed to accept hazardous waste (subtitle Clandfills). The determination of how protective disposal in a subtitle D landfill would be willdepend on the mobility of the contaminants placed in the landfill and the constructionspecifications of the landfill. DDT and its congeners, which constitute the principal threat atthe Site, are not very mobile. DDT and its congeners prefer to stay bound to organic matterand are hydrophobia. If the landfill is constructed with bottom liners and leachate extraction,it is likely that this option would be protective of human health and the environment. Thedirect contact exposure pathway would be eliminated with this alternative.
RCRA subtitle C landfills are constructed for the purpose of disposal of industrialhazardous waste. Construction requirements for bottom liners, leachate collection andmonitoring are more stringent than for landfills designed to accept municipal waste (subtitle Dlandfills). The determination of how protective disposal in a subtitle C landfill is will dependon the mobility of the contaminants placed in the landfill and the construction specifications ofthe landfill. DDT and its congeners, which constitute the principal threat at the Site, are notvery mobile. DDT and its congeners prefer to stay bound to organic matter and arehydrophobic. If the landfill is constructed with bottom liners and leachate extraction, it islikely that this option would be protective of human health and the environment. The directcontact exposure pathway would be eliminated with this alternative.
26
Compliance with ARARsExcavation and/or dredging activities would comply with the substantive requirements
of the Federal Clean Water Act and the Michigan Inland Lakes and Streams Act and Statechemical-specific, action-specific, and location-specific ARARs specified in the RI/FS Report.Contaminated sediments were tested by the TCLP (Toxicity Characteristic LeachingProcedure) and determined not to be RCRA characteristic. U.S. EPA will continuouslycharacterize the stockpiled sediments prior to disposal, typically every 200 cubic yards. Thecontaminated sediments are not considered to be RCRA listed waste because the contaminationoccurred primarily from the direct discharge of DDT process waste waters to the Pine River.See 40 CFR Section 261.33(d) comment. Waters from the dewatering process will be treatedusing Best Available- Technology (BAT) processes in order to comply with the substantiverequirements under the Clean Water Act and Part 31, Water Resources Protection, of theMichigan Natural Resources and Environmental Protection Act, 1994 PA 451, as amended(NREPA)
Excavation and/or mechanical dredging would comply with action- and location-specific ARARs, through appropriate management (i.e., storage and disposal) of dredgedmaterials.
During implementation of this alternative reasonable efforts will be made to minimizepotential human health and ecological risks, as well as minimize disruption of existing naturalprocesses associated with remediation. Excavation and/or mechanical dredging is not expectedto create an unacceptable short-term risk to human health or the environment fromresuspension of contaminated sediments based on two sediment excavation projects U.S. EPAhas completed. Engineering controls, such as silt curtains and temporary coffer dams wouldbe utilized to minimize concerns with resuspension in conjunction with turbidity monitoring.
Transportation of the waste material will maintain the proper permits and adhere tothe applicable standards for proper identification number, reporting and manifest system asestablished by the U.S. and Michigan Department's of Transportation. This alternative wouldmeet the chemical-specific, action-specific and location-specific requirements set forth infederal and state laws.
Cost EffectivenessIn the Agency's judgment, the Selected Remedy is cost-effective and represents a
reasonable value f^r the money to be spent. In making this determination, the followingdefinition was used: "A remedy shall be cost-effective if its costs are proportional to its overalleffectiveness." (40 CFR 300.430(f)(l)(ii)(D). This was accomplished by evaluating the"overall effectiveness" of those alternatives that satisfied the threshold criteria (i.e., were bothprotective of human health and the environment and ARAR-compliant). Overall effectivenesswas evaluated by assessing three of the five balancing criteria in combination (long-termeffectiveness and permanence; reduction in toxicity, mobility, and volume through treatment;and short-term effectiveness). Overall effectiveness was then compared to costs to determinecost-effectiveness. The relationship of the overall effectiveness of this remedial alternative wasdetermined to be proportional to its costs and hence represent a reasonable value for the moneyto be spent.
27
For this Site, Alternatives 2A, 3A, 4 and 7 were all considered to be cost-effective.The cost of Alternative 7 (capping), although significantly lower than costs for Alternatives2A, 3A and 4 (see Cost Table on page 23 of this ROD), was not considered to be proportionalto overall effectiveness. The long-term effectiveness and permanence of a shallow water cap isuncertain since capping in shallow waters has never been completed. In addition, even at siteswhere capping was implemented, none of these sites have post-monitoring data to show thatthey have been effective at reducing contaminant levels in fish. Alternative 7 does not employany treatment and therefore will not reduce toxicity, mobility, and volume through treatment,and does not provide any additional degree of short-term protectiveness over Alternatives 2A,3A and 4 since installation of a cap will result in some resuspension of contaminated sedimentsjust as dredging or excavation will.
Alternatives 2A, 3A and 4 are very similar and have similar costs (see Cost Table onpage 23 of this ROD). In evaluating the incremental cost-effectiveness of these alternatives,the decisive factors considered were amount of water required to be treated and efficiency atremoving sediment. Because Alternative 4 utilizes temporary coffer dams and excavation ofsediment, it is anticipated that sediment removal wili occur faster than using mechanical orhydraulic dredging and the amount of water that will require treatment is less than for either ofthe dredging alternatives. Therefore the cost for Alternative 4 is lower ($16.9 million) andrepresents a better value for the money spent. The cost for the Selected Remedies,Alternatives 4, 5 and 6 ranges from $20.1 - 34.3 million depending on how much of thesediment is disposed in a subtitle C versus subtitle D landfill.
Utilization of Permanent Solutions and Alternative Treatment Technologies to the MaximumExtent Practicable
The Selected Remedy represents the maximum extent to which permanent solutions andtreatment are practicable at the Site and affords the best balance of tradeoffs as compared to theother alternatives. Alternatives 2A, 3A and 4 are all very simi'^r. The advantage Alternative4 has over Alternatives 2A and 3A is that with the use of temporary coffer dams the sedimentscan be excavated which will be faster than dredging and require less water treatment, therebyreducing costs. Alternative 7, capping, does not provide the best balance of tradeoffs becausethere is too much uncertainty that capping will provide long-term effectiveness and permanenceand reduction of toxicity, mobility, or volume through treatment. Implementablity is also aproblem in the St. Louis Impoundment because water levels are very shallow. The communitydoes not support leaving contaminated sediments in place.
Preference for Treatment as a Principal Element
Remedial actions in which treatment which permanently and significantly reduces thevolume, toxicity or mobility of the hazardous substances, pollutants, and contaminants is aprincipal element, are to be preferred over remedial actions not involving such treatment.
Alternatives 2A, 3A and 4 all employ treatment of the principle threat (DDT-contaminated
28
sediments) with a drying/stabilizing agent. The drying/stabilizing agent will reduce mobility ofthe DDT in the sediment prior to the sediment being disposed off-Site. Alternative 7, capping,does not employ treatment of sediments and therefore is not preferred over the selected remedy,Alternative 4.
Since the contaminated sediments are not a listed or characteristic waste under RCRA, asdiscussed in the "Compliance with ARARs" section above, the land disposal restriction forDDT-contaminated soil would not be applicable.
Five-Year Review Requirem:nts
CERCLA section 121(c) requires a five-year review if the remedial action results inhazardous substances remaining on-site above levels that allow for unlimited use andunrestricted exposure. All sediments at and above 5 ppm total DDT will be removed and it isanticipated there will not be site-related use restrictions on the river after the completion of theremedy. Restrictions on fish consumption, however, will remain in place until it isdemonstrated that contaminant levels in fish have reached acceptable levels. Therefore five-year reviews will be required until fish tissue levels reach acceptable levels. The five yearreviews for OU2 will be included as part of the five-year reviews required for OU1.
Therefore five-year reviews will be required until analyses indicate acceptable levels areachieved in fish tissue. The five year reviews for OU2 will be included as part of the five-yearreviews required for OU1.
XIII. Documentation of Significant Changes
The Proposed Plan, issued on September 8, 1998 recommended Alternatives 4, 5 and6. U.S. EPA has made a minor change in that this ROD recognizes that there may be parts ofthe St. Louis Impoundment that will require hydraulic or mechanical dredging. The RODallows U.S. EPA flexibility to use dredging in areas where excavation may not beimplementable.
29
APPENDIX A
Velsicol Chemical Corporation (Pine River) Superfund SiteSt. Louis, Michican
Figure 2-1
Velsicol ChemicalCompany
Former Plant Site
Figure 2-2
Pine River, Chippewa River, Tittabawassee River, Saginaw River and Bay
StalAHwv 61
/' Estey RdJ
^45
" • L.
u
»•o«•taCz;
E
i H*^
^inconr
/TittabawasFliver
StreetMap USA ^/\/ Highway/\/ Primary road
A\ Secondary roadWater bodyPark "'
Velsicol Chemical CorporationSuperfund SiteSt. Louis, Michigan __
Figure 23
Interpolation of Maximuni Total DDT Concentration
Total DDT ConcentrationsGreater than 10ppmArea: 31 acresVolume: 260,330 cy
Veisicol ChemicalCorporation
Former Plant Site
U.S ERAFfcgtonS2/OW96
rSglcwrl d'toWcolW plnwiv ajx
Maximum Concentration, ppm0-4.995-49.9950-99.99100-299.99300 - 499.99500 - 999.991000-4999.995000 - 9999.9910000-33000
Velsicol Chemical CorporationSuperfund SiteSt. Louis. Michigan
Figure 7.-A
Interpolation of Maximum Total DOT ConcentrationTotal DOT ConcentrationsGreater than 100 ppmArea: 20 acresVolume: 170,000 cy
Velsicol ChemicalCorporation
Former Plant Site)
US EPARegions2/oo/ee
rSgtnwl d \veMcoT£7 plnariv apr
Maximum Concentration, ppm0-4.995-49.9950-99.99100-299.99300-499.99500-999.991000-4999.995000 - 9999.9910000-33000 N
Velsicol Chemical CorporationSuperfund SiteSt. Louis. Michiaan
Mgure
Interpolation of Maximum Total DOT Concentration
Total DOT ConcentrationsGreater than 500 ppmArea: 12 acresVolume: 100,000 cy
Velsicol ChemicalCorporation
Former Plant Site
Maximum Concentration, ppm0-4.995-49.9950 - 99.99100 - 299.99300 - 499.99500 - 999.991000-4999.995000 - 9999.9910000-33000 N
jiw>«rv^ -fc ^, US ERA
^f R^onS^ 2fl9«e
r50s*n cM*toicol\87 FIELDS
Velsicol Chemical CorporatioiSuperfund SiteSt. Louis, Michigan
Hgure 2-b
Interpolation of Maximum Total DDT ConcentrationTotal DDT ConcentrationsGreater than 1000 ppmArea: 6 acresVolume: 48,000 cy
Velsicol ChemicalCorporation
Former Plant Site
Maximum Concentration, ppm0-4.995-49.9950-99.99
__100-299.99
E 300 - 499.99500 - 999.991000-4999.99
Bi 5000 - 9999.99•110000-33000
Velsicol Chemical Corporation m >/ne River) Superfund LSt. Louis, Michigan
, •>.,. .. *
Figure 2-''O
1997 USER A, GLNPO& MDEQ SedimentSampling Survey
In July 1997, the USEPA,Region 5 GLNPO and theMDEQ conducted a secondsampling in the Pine Riverand the Impoundmentarea. The survey wasintended to supplementthe May 1996 surveyand provide additionalinformation as to ihe »nature and extent of theDOT contamination inthe Pine River. A iotalDDT maximum of32,600 ppm wasdetected in one sampleat depth interval6 to 42 inches.
N
1997 ctott (USEPA. GLNPO, MDEQ)Vdftlcol CtMmlcal Corporation
US EPAFjpon S2JOUSK FIK1.HS
APPENDIX B
APPENDIX BCOST ANALYSIS FOR SEDIMENT REMOVAL ACTION ALTERNATIVE
3 PAGES
REDACTED
NOT RELEVANT TO THE SELECTION OF THE REMOVAL ACTION
APPENDIX C
Responsiveness SummaryVelsicol Chemical Superfund Site, St. Louis, Michigan
Record of Decision
I. OVERVIEW
The public participation requirements of CERCLA §113(k)(2)(D)(i-v) and CERCLA §117 havebeen met during the Record of Decision (ROD) process for the Velsicol Chemical Superfundsite ("Site"). Sections 113(k)(2)(B)(iv) and 117(b) of CERCLA require U.S. EPA to respond".. .to each of the significant comments, criticisms, and new data submitted in written or oralpresentations" on a proposed plan for a remedial action. This Responsiveness Summaryaddresses those concerns expressed by the public, potentially responsible parties (PRPs), an*}governmental bodies in written and oral comments received by U.S. EPA regarding theproposed plan.
II. BACKGROUND ON COMMUNITY INVOLVEMENT
U.S. EPA issued a fact sheet/proposed plan in September 1998 prior to the start of the publiccomment period. A public notice summarizing the Proposed Plan and announcing the publiccomment period and public meeting was published in the September 9 and 13 editions of theAlma Morning Sun, a local paper with distribution in the communities surrounding the site. Inaddition, the Proposed Plan was directly mailed to approximately 240 addressees on the VelsicolChemical mailing list. Personal press contacts with the Alma Morning Sun and Saginaw Newsresulted in news stories in the editorial columns of both publications prior to the meeting.
An information repository was maintained at the T.A. Cutler Memorial Library, 110 WestSaginaw Street, St. Louis, Michigan. Administrative Records were made available to the publicfor review at the U.S. EPA Region 5 office in Chicago and at the T.A. Cutler Memorial Libraryin St. Louis.
Community interviews were conducted in St. Louis and Alma, MI between December 15-17,1997 to provide information for U.S. EPA's Community Involvement Plan (CIP). Thoseinterviewed included citizens, city officials and members of local environmental groups. TheCIP was completed and placed in the repository and administrative records in March, 1998.
The public comment period ran from September 8, 1998, to October 8, 1998. U.S. EPAreceived no requests for an extension to the 30-day comment period.
U.S. EPA hosted a public meeting in St. Louis on September 16, 1998, to provide backgroundinformation on the Site, to provide details of the proposed cleanup plan, and to take oral publiccomments. U.S. EPA answered questions about the Site and the proposed cleanup alternativeunder consideration. The entire public meeting including formal oral comments on the
•1
Proposed Plan were documented by a court reporter. A verbatim transcript of this publicmeeting has been placed in the information repository and in the Administrative Record.Written comments were also accepted at this meeting. The meeting was attended byapproximately 100 persons, including local residents.
During the comment period, U.S. EPA received written comments from 7 commenters and 5oral comments concerning the proposed plan. Comments received during the public commentperiod and the U.S. EPA's responses to those comments are included in this ResponsivenessSummary which is a part of the Record of Decision.
HI. SUMMARY OF COMMENTS RECEIVED DURING PUBLIC COMMENT PERIOD
U.S. EPA received 5 oral comments at the public meeting which are addressed below.
1. Comment: Mr. Mike Bazda asked about the effects that dredging might have on airpollution and wanted to know if air pollution considerations were taken into accountwhen deciding between Alternatives 2A, 3A and 4?
Response: EPA's response action will comply with the Michigan Air Pollution Act. Airpollution is not expected to be a significant exposure route for the community because thesediments will be damp during handling thus reducing the likelihood of blowing. Air monitoringwill be conducted during sediment removal to verify the state's air pollution requirements aremet.
2. Comment: Mr. Dennis Busque asked if EPA had considered killing all the contaminatedfish.
Response: Killing the contaminated fish is not part of the proposed cleanup plan for the PineRiver. EPA did not consider this during the RI/FS process. If the citizens of St. Louis areinterested in pursuing this they should contact the Michigan Department of EnvironmentalQuality.
3. Comment: Mr. Harry Vitek stated that he supports the removal of the sediments but feelsEPA has not addressed the problem of scouring outside the containment curtain. He feelsEPA has no contingency plan for containing scoured sediments from moving down river.
Response: EPA does not believe that the current sheet pile configuration causes any significantalteration in the river flow or the speed of flow. Therefore, no additional scouring from ouractivities should result. During the fall EPA conducted turbidity monitoring 3 times a day toverify that resuspension was not occurring, it wasn't. Turbidity monitoring will continue byEPA every day that EPA is on the river doing work. In the upcoming years of work, if there is apotential for the cofferdam configuration to increase flow velocity, and therefore scour, EPA willconsider installing a catch basin below the St. Louis dam to collect any resuspended sediments
before they move downriver.
4. Comment: Dr. Ed. Lorenz, Community Advisory Group (CAG) Chairman, commentedthat the CAG unanimously supports the removal of the contaminated sediments from thePine River.
Response: EPA appreciates the support of the CAG.
5. Comment: Ms. Tara Blanford wanted to know if the kiln dust that will be mixed v-ith thesediment to dry it out is more toxic than DDT and if it will blow when EPA is mixing it?
Response: Kiln dust would only be toxic if it contained contaminants such as heavy metals.EPA will ensure that the kiln dust is not toxic by having it sampled before it is shipped to thesite. During the mixing operations it is likely seme kiln dust will blow off site. EPA will betaking precautions to keep blowing to a minimum, such as housing the kiln dust in a silo andusing a tremie pipe or similar application method to apply it to the sediments.
U.S. EPA received comments from seven commenters during the public comment period whichare addressed below.
6. Dr. Eugene Kenaga submitted seven comments that were printed in the October 4, 1998Midland Daily News. EPA saw the article and responded to Dr. Kenaga with a letter.Dr. Kenaga clipped the article and submitted it to EPA during the public comment period.EPA's response is re-printed below in its entirety and summarizes Dr. Kenaga'scomments.
October 16, 1998
Dr. Eugene KenagaPresident, Little Forks Conservancy1584 E. Pine River RoadMidland, Michigan 48640
Deai Dr. Kenaga,
A co-worker of mine has family that lives in Midland. He passed on to me the Sunday, October4, 1998 Midland Daily News FORUM column entitled Questions on Pine River cleanup needanswers (see attached). I am the U.S. EPA project manager for the Velsicol site in St. Louis,Michigan which includes cleanup of the Pine River. I am happy to answer your questions posedin the Midland Daily news and I encourage you to call me if you have any further questions at 1-800-621-8431 extension 3-6576 or direct dial at (312)353-6576. I am also available to meet withyou and any other concerned citizens about the project. You are also invited to attend monthlyCommunity Advisory Group (CAG) meetings on the third Wednesday of each month in St.
Louis, Please contact Dr. Ed Lorenz for more information about the CAG at (517) 463-7203.
For the sake of brevity, I have transcribed segments of your questions, the full text is attached tothe end of this letter. Here are my responses to your questions:
Question 1: "Why, after 20 plus years of little or no action are EPA and DEQ trying to clean upthis river site?"
Response: EPA and the State of Michigan (MDEQ) settled Velsicol's liability for the mainplant site and contamination in the Pine River in 1982. The 1982 agreement required Velsicol toclose the main plant site by constructing a slurry wall around the entire 50 acre former facilityand place a clay cap on top. The purpose of this was to interrupt contamination migration fromthe site to the Pine River. Velsicol was not required to do any cleanup work on the river in 19H,and received a covenant not to sue from U.S. EPA and the State. This means that Velsicol wou'dhave no further legal obligation to clean up the river. When this agreement was made, in 1982,dredging of hazardous waste was not commonly done, and hazardous waste dredgingtechnologies were not fully developed. Today, hazardous waste dredging is common, and thetechnology necessary to ensure safe dredging operations has been developed. So, in 1995, whenMDEQ fish monitoring discovered that contaminant levels in Pine River fish had doubled sincethe last time they collected fish in 1989, the Agencies became concerned that additionalcontamination was entering the Pine River from the main plant site. In 1996, Velsicol re-assessed the slurry wall and clay cap constructed on the main plant site to see if it was the sourceof additional contamination to the Pine River. The results showed that the slurry wall iseffectively keeping any additional contamination from reaching the Pine River, however Velsicoldiscovered that the clay cap is leaking, allowing rain water to enter the slurry wall and capsystem. This accumulated rain water is pumped out routinely by Velsicol and taken to apretreatment facility near Detroit.
At the same time Velsicol was re-assessing the cap and slurry wall, U.S. EPA and MDEQdecided to take more sediment samples from the Pine River, to see if contaminant levels insediments have changed since 1982. The results showed that contaminant levels in the top 6inches have decreased since 1982, and contaminant levels below 6 inches are similar to 1982le\ .Is, very high. More sediment samples and more fish tissue samples were collected by U.S.EPA and MDEQ in 1997. This data confirmed that levels in fish were still higher than in 1989.Our conclusion was that although there did not appear to be any new sources of contamination tothe Pine River, the fish were getting more contamination and storing it in their tissues. Why thisis happening is explained in EPA's response to Comment #12-5.
So EPA decided it should re-consider the decision in 1982 to leave contaminated sediments inthe Pine River. With hazardous waste dredging technologies available there was no reason not togo forward with a cleanup. Therefore, U.S. EPA completed a thorough study of fish andsediment contamination in the Pine River and evaluated cleanup alternatives in a report called theFinal Streamlined Remedial Investigation/Feasibility Study Report dated August, 1998. Several
copies are located at the T.A. Cutler Memorial Librai y in St. Louis for anyone to review. U.S.EPA proposed a cleanup that would remove all sediments above 5 ppm total DOT from the PineRiver. A public meeting was held in St. Louis to discuss the proposed plan on September 16,1998.
The 1997 sediment sampling confirmed that there was a "hot spot," an area where contaminantconcentrations exceeded 3,000 ppm total DOT. For this area U.S. EPA proposed an emergencyremoval action, which allowed U.S. EPA to start work faster when levels of contamination aredocumented to be extremely high. This work began on June 8, 1998.
All of this work was coordinated with the community through the Community Advisory Group(CAG) which meets once a month in St. Louis. The CAG has a representative member fromMidland.
Question 2: "Nothing was said at the meeting about residues at the site concerning otherchemicals such as PBB, possibly more persistent than DDT and DDE, still in the Pine River, orof HBB and TRIS made by Michigan Chemical Co. in the past."
Response: U.S. EPA and MDEQ focused on contamination in fish tissue to guide us in settingcleanup levels for sediments. This is because risk at the site is associated with eatingcontaminated fish. Our goal is to clean up fish, to do this EPA has to remove the sediments.Chemicals detected in fish in 1997 above the detection limit were mercury, DDT and itsmetabolites, chlordane congeners, PBB, hexachlorobenzene and octachlorostyrene. The DDTconcentration in most of the fish collected exceeded the Michigan Department of CommunityHealth (MDCH) Level of Concern of 5 ppm total DDT (maximum level in a carp filet was 90ppm). PBB was not detected in carp, but was found at low levels in smallmouth bass (maximumconcentration 0.3 ppm). The mercury and chlordane concentrations did not exceed the MDCHLevels of Concern. MDCH has no official Level of Concern for the other chemicals detected.So the chemical of most concern found in fish was DDT and its metabolites (DDD and DDE).Therefore EPA focused on total DDT in sediments.
In addition, U.S. EPA's analysis of sediment data shows that to achieve the cleanup goal of 5ppm total DDT in the area to be dredged all sediments would need to be removed (approximately260,000 cubic yards). This would require an estimated average dredging depth of 5 feet to theriver bottom, which is clay. Since all sediment will be removed, all other contaminants ofconcern would be removed also. The clay will then be sampled to ensure the cleanup goal isattained. In addition to total DDT, the confirmation samples will also be. analyzed for PBB, HBBand TRIS.
Question 3: "The company that did EPA's dredging of the Fox River at Manistique (WI) may bedoing the dredging of the Pine River. Is it? According to the Fox River Group, PCB's increased5-fold in the sediments after dredging at Manistique. The cost over-run there ballooned to over$22 million after an original estimate of $4-6 million and is still not done. Is this same thing
going to happen because of the Pine River dredging "remedial" action?"
Response: For clarification, EPA is currently hydraulically dredging the ManistiqueRiver/Harbor in Manistique, Michigan. The Fox River is also a contaminated sediment site EPAis considering cleaning up located in Appleton/Green Bay, Wisconsin. The company that isdoing the dredging at the Manistique River/Harbor site will not be doing the sediment removal atthe Pine River site in St. Louis, Michigan. Regarding the dredging project at Manistique,Michigan, U.S. EPA considers this project to be a success. U.S. EPA has documented a 40-foldreduction of PCB contaminant concentrations in areas where the dredging is complete.
The information from the Fox River Group regarding increases in PCB contamination duringdredging refers to sampling of areas that were not completely dredged by EPA. However, asstated above, areas that have been completely dredged by U.S. EPA have shown a 40-foldreduction of PCB concentrations.
U.S. EPA's Action Memorandum dated September 10,1996 estimates the cost of the dredging atManistique River/Harbor to be $15 million. U.S. EPA had estimated a partial cleanup of theRiver/Harbor to be $6 million. Currently U.S. EPA is estimating the total cost of the project willbe about $30 million and will be complete in the fall, 1999. The increase in the cost estimatewas due to unknown sediment conditions, such as significant amounts of rock and slab woodwhich slowed down the rate that contaminated sediments could be removed.
It is possible that the cost estimate for the Pine River sediment removal project will be greater orlesser than the estimated $20 - 34 million dollars. Cost estimates are just that, estimates. Actualsite conditions can significantly alter cost estimates.
Question 4: "What assurances do we have that melt from snow or other floods will not overruncofferdams within the remedial sites. Will you be able to treat the water successfully or have thecapacity when the flow increases or sediment increases?"
Response: The cofferdams U.S. EPA is installing are made of sheet pile. It is very sturdy andinterlocks to form water tight seals. The sheet pile is driven down through the sediment ten feetinto the clay bottom of the river and is braced on the inside and outside for additional support. Itis very unlikely '.it snow melt or other floods would overrun the sheet pile. If there was a floodand the containment area was flooded then EPA would shut down operations until flood waterssubsided outside of the cofferdam. Flood waters remaining inside the cofferdam would bepumped to the treatment plant EPA will build on Velsicol's property and treated to Statestandards before being discharged to the Pine River.
Question 5: "The EPA judgment of successful remediation seems to come from the amount ofmaterial removed from the Superftmd Site, not upon how much pollution and disturbances willresult down river. Is this the only correct way to assess cleanup?"
Response: U.S. EPA and MDEQ do not expect there to be any significant pollution ordisturbances down river from this sediment removal project. A silt curtain has been installedaround the sheet pile. Turbidity monitoring in the river is conducted several times a day toensure that no significant releases of sediments occur downriver. The no consumption advisoryfor all species offish is in effect for the portion of the Pine River that flows through Midland.Cleaning up the contaminated sediments in St. Louis is expected to result in fish that willsomeday be clean enough to eat. This will benefit Midland as well as St. Louis and all othertowns on the Pine River that are subject to the fish advisory. The cleanup benefits to the fish andall people and animals that live alonp the river far outweighs the risks from the small amount ofre-suspended sediment. To leave the contaminated sediments in piace and vulnerable to naturaldisasters pose more risks to the St. Louis and downstreair communities.
Question 6: MDEQ has a warning not to eat fish from the Pine River. There is evidence that thePine River at Midland has recovered its invertebrate and fish populations now many years afterthe Michigan Chemical Co. episode. What assurances do we have that the persistent sediment orsoluble residue in the remedial action that could come down the Pine River will not reverse thisrecovery?
Response: This questions is similar to Question 5. The only thing I would add is that DOT,ODD and DDE are hydrophobic, meaning they do not solubilize easily in water. In fact, watersamples taken in the St. Louis Impoundment show no detection of DDT, ODD or DDE. Thesechemicals prefer to stay in the sediments and will generally not partition into water.
Question 7:1 pump water from the Pine River to water my lawn and garden What assurance do Ihave that eating these vegetables is safe or that I will not build up toxic residues in the soil fromthis practice?
Response: This question is similar to Question 6. There are no detectable levels ofcontaminants in water from the Pine River. Therefore irrigating your lawn and crops is safe. Ifyou would like to discuss this concern further with the Michigan Department of CommunityHealth (MDCH) you may contact Brendan Boyle at (517)335-8138.
In closing, the proposed sediment removal project has received approval from U.S. EPA, MDEQ,Micnigan Department of Community Health (MDCH), and the Community Advisory Group(CAG). U.S. EPA consults continuously with these groups on the progress being made at thesite. U.S. EPA is taking every reasonable precaution to protect our employees, contractors, thecommunity and downriver communities from exposure to contaminated sediments during theproject. I encourage you to call me directly with any other questions or concerns you may haveregarding this project.
Sincerely,
Beth ReinerRemedial Project Manager
cc: Kim Sakowski, MDEQBrendan Boyle, MDCHEd Lorenz, CAG ChairmanSam Borries, EPA OSCStuart Hill, EPA CICMidland Daily News
7. Mr. Frederick L. Brown submitted 5 comments as follows:
Comment #7-1: No explanation for the significant increase of DOT in fisli has beenoffered. Is there some other source of contamination to the river?
Response: Pages 24-27 of the RI/FS summarizes the fish data and discusses possible reasons forwhy the concentrations in fish after 1989 are significantly higher than pre-1989. Fish tissue datais difficult to compare due to variabilities in the data (weight, age, %fat and number of samplescollected). The average weight offish collected after 1989 increased, and sometimes so did theaverage %fat. This would account for some of the reason fish tissue levels of total DOT aregreater post-1989, since larger, older fish will bioaccumulate more DDT. Also, the 1997 dataanalyzed for all 6 congeners of total DDT, all previous years had only analyzed for 3 congeners(see Table 2.2-10, page 25 in the RI/FS Report). EPA's conclusion is that if fish tissue datacould be normalized it would show that levels have been fairly constant over time.
When the 1994 fish tissue results came out, both EPA and MDEQ were concerned that the mainplant site might be leaking additional contamination to the Pine River. Velsicol conducted athorough study of the slurry wall and cap system with EPA/MDEQ oversight and found that thecap was leaking but the wall appears to be intact. Therefore, EPA and MDEQ concluded that themain plant site is not leaking additional contamination into the Pine River. MDEQ also wentthrough its records looking for other industries or disposal sites along the river that may havecontributed DDT to the river. None were found. So the answer is no, EPA and MDEQ have notfound any additional sources of DDT contamination to the Pine River in the area of the Velsicolsite.
Comment #7-2: What about contaminants other than DDT?
Response: A similar question was asked by Dr. Kenaga, please see our response to his Question#2.
Comment #7-3: What about the potential for downstream releases? Construction of
coffer dams will restrict the river channel, increase flow velocity and cause scouring andresuspension. This will also be affected by spring run-off and/or high rain fall.
Response: The sheet pile which is currently in the river does not obstruct the rivers naturalchannel and therefore there should be no increase in flow or scour. However, EPA is takingprecautions to ensure that significant releases of resuspended sediments do not move down river.A silt curtain has been installed around the sheet pile. During construction activities, turbiditymonitoring in the river is conducted three times a day to verify that no significant releases ofsediments occur downriver. EPA does not believe that the current sheet pile configurationcauses any significant alteration in the river flow or the speed of flow. Therefore, no additionalscouring from our activities should result. When the remedial portion of the work begins thecoffer dams may be reconfigured and if the reconfiguration may increase flow velocity, andtherefore scour, EPA will consider installing a catch basin below the St. Louis dam to collect anyresuspended sediments before they move downriver.
Comment #7-4: Monitoring of impacts of downstream releases. Deposition of releasedsediment should be mapped, water column concentration and wildlife bio-uptake of allcontaminants should be determined and large scale agricultural irrigation should bemonitored. Downstream aquatic population dynamics should be determined before andafter dredging.
Response: In 1996 MDEQ collected sediment samples between the St. Louis dam and theconfluence with the Chippewa River. MDEQ had difficulty finding locations below the St.Louis dam where there was enough sediment deposition to obtain a sample. Since there aren'tmany depositional areas below the dam it would be difficult to map deposition of any releasedsediment. In addition, turbidity monitoring completed by EPA during the fall of 1998 indicatedthat turbidity levels did not increase during construction activities such as installation of the siltcurtain and sheet pile. Data indicates that the main exposure route to wildlife fromcontamination in the Pine River is from fish consumption. Therefore, addressing the source ofcontamination to fish should reduce the bio-uptake of contaminants to wildlife. If you wouldlike to discuss this concern further with the Michigan Department of Community Health(MDCH) you may contact Brendan Boyle at (517)335-8138.
Regarding irrigation, the Pine River fish monitoring data show that downstream lish haveaccumulated significant concentrations of DDTr (sum of DDT, DDD and DDE) prior to clean upactivities. Since depositional zones are scarce below the St. Louis dam, the primary route ofexposure to the downstream fish is probably direct uptake from the water column. Based on the1997 carp fillet data, the calculated downstream water concentration is approximately 0.2 ppbpp-DDTr (sum of the para, para-DDTr - see EPA's response to Comment #12-5 for calculations).In terms of irrigation use, this is equivalent to 20.6 mg pp-DDTr/acre-inch irrigation water (0.2ug/1 x 102,790 I/acre-inch). Incorporated into the top 6 inches of soil, this results in 28 ppt pp-DDTr in soil for every inch of irrigation water applied (20.6 mg/acre + 746,500 kg/acre-furrowslice). For a quick benchmark comparison, the soil level of concern for DDT in The Netherlands
is 4 ppm. 4 ppm is 143,000 times greater than 28 ppt. Farmers would have to apply 12,000 feetof irrigation water before their fields would accumulate DDT to the Netherlands level of concern(this would be more than 2 miles of irrigation water). Note: this calculation does not account fordegradation or volatilization of DDT after it is in the soil, which would further increase theamount of irrigation water needed to reach soil levels of concern.
Even low levels of DDT in agricultural soil could be of concern if plants bioconcentrate DDTinto their tissues to levels much greater than in soil (as occurs between fish and surface water).However, this is not the case. Although plants can absorb DDT from the soil, the levels in plantsremain low. For example, soybeans and ryegrass accumulate DDT in their above-ground oarts toconcentrations only 14 to 93 times higher, respectively, th?n in the solutions bathing the roots(cited in Polder, et a!. 1995), compared with 1.5 million times greater concentrations in fishlipids compared with surface water. Applying these relationships to irrigation water with 0.2 ppbDDTr, crops would accumulate 3 to 19 ppb DDTr in the above-ground tissues (0.2 x 14 or 93).As a rough comparison, this is less than 4 % of the PDA limit of 5 ppm. Even for a worse casescenario in which downstream surface water concentrations reach the solubility limit of DDT of1.2 ppb (Matsumura 1985), the PDA level would not be exceeded by crops grown entirely withriver water (17 ppb to 1 ppm).
In conclusion, the risks associated with crops grown with Pine River irrigation water are minor.Monitoring efforts should therefore focus on the much greater potential risks associated withriver fish.
The suggestion to monitor aquatic population dynamics does not specify the type of organism.The logical choices would be aquatic plants (macrophytes and/or algae), benthic invertebrates, orfish. Each is discussed below in terms of susceptibility and use of the monitoring data fordecision-making purposes.
Aquatic plants would not be suitable because the toxicity of DDTr to plants is very low. Also,plants are not an important exposure pathway for accumulation of DDTr by other aquaticorganisms (Matsumura 1985). They are not considered further.
Most of the benthic (bottom-dwelling) invertebrates are insects, so they are expected to besusceptible to DDTr since it was marketed as an insecticide (e.g., Edwards, et al. 1964); however,several studies hu\ e shown evolution of resistance to DDTr by benthic invertebrates inchronically contaminated waterways (Webber, et al. 1989 and refs.). Since downstream benthiccommunities have been chronically exposed for several decades to surface water DDTr dissolvedfrom the St. Louis impoundment, they probably have evolved higher levels of resistancecompared with unexposed communities. Downstream benthic communities may be unlikely torespond sensitively to changes in surface water DDTr in the Pine River. Another complication isthat base-line benthic population/community data are lacking for the Pine River. This isimportant because benthic measurements vary by season and by year. A single set of datacollected in the spring before the beginning of the next construction season would not provide a
10
suitable base-line for evaluating benthic changes during remedial activities. There will be noreliable methods for determining whether observed trends in the benthic surveys are related tothe site activities (increased contaminant leading and/or turbidity), or due to natural seasonal orannual fluctuations (including effects of weather). The only reliable determination will be in theevent that downstream benthos is catastrophically impacted, that is, if the downstreamcommunity is virtually eliminated. Trends other than catastrophic will likely not be attributablesolely to remedial activities, and so will be of limited use for decision-making. Benthicmonitoring is therefore recommended only with the understanding that it will provide limitedinformation - i.e., whether catastrophic impacts on downstream communities have occurred. Itwill be unreliable for less severe impacts because of the lack of appropriate long-term baselinedata and the possible evolution of DOT resistance by benthos over the past several decades ofchronic exposure. EPA and MDEQ are discussing conducting some benthic monitoring but nodecision has been made yet.
Fish also are unlikely to be poisoned outright by DOT releases, as evidenced by the survival ofcarp and other species within the St. Louis impoundment. Monitoring fish populations willtherefore not be an appropriately sensitive endpoint for detecting adverse downstream impacts.
The recommended approach for monitoring for downstream releases is to continue the fishcontaminant surveys performed by the state. This has several advantages. An important one isthat historical baseline data exist, so there is a good foundation for comparison with the remedialmonitoring data. Another advantage is it measures the parameter most directly associated withboth human health and ecological risks - the concentration of DDTr in fish. EPA recommendscollection of carp and analysis of fillets so that the data is directly comparable to previousmonitoring. The data should be lipid-normalized for the sum of pp-DDT, pp-DDD and pp-DDE.Since downstream fish appear to be exposed primarily through direct water column partitioning,this should provide a sensitive measure of possible releases from the site to downstream reaches.
Comment #7-5: The cleanup standard of 5 ppm total DOT in sediments is based on riskto adults via fish consumption and reproductive problems in fish eating birds. Possibleendocrine and neurological impacts were mentioned but no data presented, why? Failureto consider exposure to children and fetal dose caused by exposure via mother's bodyburden of contaminants and breast feeding were not considered and should have been.
Response: As was stated in the RI/FS report, the toxicity information needed to look atendocrine and neurological impacts is not complete. It is certainly agreed that fetal exposuresand maternal body burden are important considerations, and it would have been preferable toassess these endpoints. However, no toxicity factors for these endpoints exist at this time. EPAtherefore looked at all endpoints EPA could, i.e. those that are in IRIS. Short of deriving a new,non-peer reviewed toxicity factor, this type of assessment is not possible at this time.
Risk assessments should be performed with defensible data. Defensible data have beendeveloped for assessing the reproductive effects of DOT in fish-eating birds (which is a type of
11
endocrine disruption effect). Although other effects, such as behavioral, have been reported insome wildlife studies, the levels associated with unacceptable risk are not well characterized.Unless the dose levels associated with unacceptable risk to wildlife are known, there is no way touse the results of food-chain modeling to guide risk management decisions. Such levels areknown for bird reproduction, but are not adequately known for behavioral or other effects inwildlife (other than direct lethality, which occurs in mammals and birds at exposures muchhigher than that at this site).
8. C.W. Dunbar submitted one comment:
Comment #8-1: Alternative 1 (No Action) should be the selected alternative. In the FS itis obvious that the purpose of the work is to remove DDT from the fish in the Pine River.Unless you remove every trace of DDT from the river, fish will continue to build up POTcontent in their bodies. It is physically impossible to remove every trace of DDT fromthe river and therefore this project is an exercise in futility.
Response: EPA calculated the levels of contamination in fish tissue if 5 ppm total DDT and 1ppm total DDT were left behind in sediments in the St. Louis Impoundment (see page 80-81 inthe RI/FS Report). These calculations indicate that if sediments were cleaned up to 5 ppm totalDDT or less carp fish tissue levels would be 1.7 ppm total DDT and smallmouth bass fish tissuelevels would be 0.8 ppm total DDT. These levels result in an acceptable risk to human healthand a significant improvement from levels currently seen in fish tissue. Although the river willnever be returned to pristine conditions, EPA believes the cost of the response action is worthreturning the Pine River to a place that is safe for humans and wildlife.
9. Mr. Rick McKenna wrote in one comment:
Comment #9-1: As a sportsman and an environmentalist I feel the EPA and MDEQ havea removal plan that is very adequate. The Velsicol site is not going to go away by itself.So I say we go ahead with this well thought out solution and get something done.
Response: Thank you for you comment supporting the cleanup plan.
1C. Mr. Bob Veenstra wrote in eleven comments:
Comment #10-1: For locations where the cofferdam is not adjacent to the formerVelsicol property, it is not clear how water and sediments will be transferred to theVelsicol property. Will construction equipment be stationed on property other than thatwhich is owned by Velsicol? If so, what type of access, security and environmetnalmonitoring will be done at this location?
Response: EPA realizes that the use of cofferdams and dry excavation may not be the best wayto remove all sediment from the St. Louis Impoundment which is why the ROD states that
12
dredging technologies, either mechanical or hydrau':c, may be used for some of the removal.The details of how water and sediments will be transferred to the Velsicol property have not beencompleted yet, this will happen during the remedial design phase of the project (after the ROD issigned). Property other than Velsicol's may be used for storage of equipment, but this will bedetermined during the remedial design. EPA also knows that designs can and often are modifiedwhen the actual field work starts. If other properties are used the same type of security, such asfencing, would be used as is used at the Velsicol property.
Comment #10-2: To place the cofferdams, plans have been discussed to lower the waterlevel in the St. Louis Impoundment which could expose contaminated sediments. Whatprecautions will be taken to address the public health exposure potentials of this action?
Response: EPA did propose lowering water levels in the St. Louis Impoundment to facilitate theinstallation of the sheet pile. This idea was presented to the Community Advisory Group (CAG).Many concerns were raised similar to the one in your comment. EPA heard the communitiesconcerns and changed the plan for installation of the sheet pile. Instead of lowering the waterlevels, EPA brought in sectional barges and actually raised the water level a little so the bargeswouldn't ground out. This method was successful and there seems to be no reason why EPAcould not use this method for any additional installation of sheet pile.
Comment #10-3: The process of excavating sediment, placing it on trucks, transporting itto a pad, mixing it with cement kiin dust, reloading it onto trucks and taking it off-site fordisposal requires a significant amount of material handling (and equipment manpower).A common rule in construction and material handling is that costs go up significantlyevery time something is picked up and moved. Other methods, which were evaluated,particularly hydraulic dredging, would require significantly less handling, manpower, andequipment to reach the same endpoint.
Response: EPA is proposing to mix the cement kiln dust in-situ, then excavate the sediment onto trucks, unload the stabilized sediment onto a staging pad for up to three days and then load onto trucks for disposal. This would not be significantly different than hydraulic dredging. Inhydraulic dredging the slurry would go to a settling basin where a polymer would be added. Theslurry would then be pumped to belt filter presses for dewatering. After pressing the sludgewould be moved to a staging pad and then loaded for disposal. EPA completed detailed costestimates for hydraulic dredging, mechanical dredging and excavation (see Appendix B of theRI/FS Report). Hydraulic dredging was estimated to cost approximately $5 million dollars morethan excavation. This is because hydraulic dredging creates the most wastewater that has to betreated before discharge back to the Pine River. Also hydraulic dredging is not estimated to beable to remove sediment as fast as excavation, therefore adding costs for additional time to rentequipment and pay contractors. However, EPA realizes that all areas of the St. LouisImpoundment may not be able to be excavated and therefore dredging, either mechanical orhydraulically, may be considered for some parts of the St. Louis Impoundment.
13
Comment #10-4; The primary technical reason that hydraulic dredging was not selectedappears to be the concern that it would generate significant amount of water to be treated.However, if the dredge were appropriately sized the water generated would be wellwithin the operational capacity of the water treatment system being designed for theemergency removal action. This concern does not appear to be valid if the dredge wereappropriately sized.
Response: Water treatment is a major concern for EPA with this project because the staterequires that EPA meet a stringent discharge standard before EPA can discharge any water backto the Pine River. EPA has completed treatability studies on the sediment and water whir ,shows that the sediment is very fine, which makes it hard to settle in a clarifier and will be amajor hurdle for meeting the discharge standard because DOT sticks to the sediment, so virtuallyall particles must be removed from the water in order for us to meet the discharge standard. Theother hurdle with hydraulic dredging is that it does not remove sediment as fast as excavation,which makes it more costly than excavation. EPA has experience with hydraulic dredging and isaware that there are pluses and minuses, as with any technology. For this particular site,excavation appears to be the most cost effective way to address contamination in the St. LouisImpoundment. Of course if excavation is not possible in some areas of the St. LouisImpoundment, dredging will be used.
Comment #10-5: The use of cement kiln dust poses several concerns, such as it may becontaminated with heavy metals, it is susceptible to wind dispersion and thus significantexposures to the local population. Appropriate monitoring programs must be establishedto ensure safe handling of the material. Also the mixing method of using the buckets ofexcavators could result in significant amounts of windborne kiln dust to be released andoffers little control over dosage or the cost effective use of this material.
Response: Cement kiln dust is a commonly used stabilizing agent. You are correct that therecan be problems with it blowing when it is stored or when it is being mixed. EPA will be takingprecautions to minimize blowing, such as storing it in a silo and applying it with a tremie pipe orsimilar application method. Also, the kiln dust EPA purchases will be tested for heavy metals toensure it is not contaminated above regulatory limits. EPA will be implementing an airmonitoring program which meets the requirements of the state of Michigan.
Commen; #10-6: In terms of short-term risks to the community and on-site workers, aswell as technical implementability, the recommended alternative appears to be lessdesirable than other alternatives reviewed.
Response: EPA disagrees. EPA's expectation is that excavation of sediments with the use ofcofferdams will remove the contaminated sediments faster and more efficiently than the dredgingtechnologies. This saves time and money and results in less likelihood of the community beingexposed, and less likelihood of significant releases moving down river. EPA does not believethat risks to the community are any greater from Alternative 4 than from the other alternatives
14
evaluated. On-site workers will use adequate personal protective equipment to protectthemselves from exposure.
Comment #10-7: The cost analysis for the hydraulic dredging alternative was developedby extrapolating costs from EPA's experience at Manistique, Michigan. The cost istherefore erroneously high because the equipment and technical approach used atManistique would be inappropriate for use at the Velsicol site. A properly designedhydraulic dredging program would be much less costly than EPA's estimate forAlternative 2A.
Response: EPA did use its experience at the Manistique, M! project to estimate costs for adredging project in the Pine River. However, costs were not just extrapolated, costs from theManistique project were used as a guideline to estimate costs for the Pine River project. Thiscertainly is appropriate and EPA does not believe our cost estimate is erroneously high. Thepurpose of the cost estimate in the RI/FS is to estimate costs within +50% to -30% of actualcosts. Therefore it truly is just a cost estimate and may be significantly different from actualcosts. The important part of the cost estimate is that all alternatives are evaluated similarly andtherefore the cost estimate gives an indication of how each alternative compares to the others.
Comment #10-8: The cost for hydraulic dredging appears to be grossly exaggeratedwhen compared to actual hydraulic dredging projects being completed today in the samegeographic region. EPA estimated divers would be necessary at a cost of $300,000 whenin reality this would rarely, if ever, be used.
Response: EPA thinks its cost estimate for Alternative 2A is reasonable based on our experiencewith the Manistique hydraulic dredging project, which is currently on-going in the state ofMichigan. Divers have been used by EPA for equipment maintenance, equipment repair,equipment retrieval, sampling if necessary, surveying/assessing if necessary and possible spotdredging around bridge piers if necessary. It is possible that divers would not be used, but it isalso possible that they would be used more than we assumed at an even greater cost.
Comment #10-9: Several of the line items in the cost for Alternative 4 do not appear tomatch the technical approach in the text of the RI/FS such as (1) there is no line item forexcavation equipment and (2) there is a line item for sectional spud barge and gravitysettling tanks.
Response: The commenter points out a valid oversight in the line item equipment charge.Alternative 4A should have two excavators listed instead of a 20-ton crane. The cost howeverwill virtually remain the same. EPA disagrees with the commenter's concern over the barge andsettling tanks. The equipment is required to complete the project. The barge is required to installthe coffer dam and the settling tanks or equivalent will be needed for primary water treatment.
Comment #10-10: It appears that the cost estimates developed in the FS are not truly
15
representative of the actual approaches that would be used on this site, but instead wereextrapolated from other jobs or are out of date and not reflective of the current thinkingon the project. As such, the relative ranking of alternative that has been conducted, inlarge part based upon the cost estimates, may not be valid.
Response: EPA disagrees that our cost estimates are out of date and not reflective of currentthinking on the project. As indicated above, the Manistique, MI project is not out of date, it iscurrently on-going. EPA used costs from the Manistique, MI project as a guideline, notverbatim. EPA's cost analysis reflects the site-specific approach envisioned for the Velsicol site.A significant amount of thought, time and energy went into the cost estimates. However,modifications to the approach typically occur after construction begins. Actual site methodsoften are not finalized until after construction begins.
Comment #10-11: Based on the technical and financial questions raised above, theselection of a recommended alternative should be re-evaluated based on correct costanalyses and a critical review of the technical implementability and health risksassociated with each alternative.
Response: EPA has modified the final cleanup plan in the ROD to include the possibility ofdredging in areas of the St. Louis Impoundment that may not be able to be excavated. EPA doesnot agree that there are significant technical or financial questions with the analysis that wascompleted in the final RI/FS report.
11. Mr. Edgar Ilgenfritz wrote the following comment:
Comment #11-1: As a downstream resident I am alarmed at the idea of scooping out thesediments and that the otherwise healthy river below the dam may not be so after thiseffort is complete. The 1986 flood scoured the river and EPA's actions may destroy thisresult. I wonder if any samples offish from below the dam have been taken.
Response: Mr. Frederick Brown made similar comments, see EPA's response Comment #7-3and #7-4 above. Fish samples have been collected below the dam and they are highlycontaminated. Removing the source of contamination from St. Louis will benefit the downriverfish and therefore the downriver residents.
12. The Velsicol Chemical Company submitted an 11-page technical report prepared byBlasland, Bouck & Lee entitled Review Summary for Final Streamlined RemedialInvestiagion/Feasibility Study Report, October 1998. Except for comments on the costestimates, it is not easily discernable which parts of the text are actual comment andwhich parts are narrative. EPA has summarized what appeared to be significantcomments below:
Comment #12-1: Section 2.1 - Site Description. The hydrology and hydraulics section
16
of the RI/FS report lack data such as basic flow statistics, velocity distributions, shearstresses during extreme flows, etc. which appears to affect the development andevaluation of the alternatives. For example, given that the alternatives include theconstruction of temporary dams, movement of sediment laden barges, or sedimentcapping more detailed analysis is required to determine if flow can be diverted or if theexisting dam can be operated to optimized barge mobility or if armoring is necessary fcucapping. No mention is made of the 1986 flood and any effect it may have had on thesystem.
Response: EPA does not agree that the hydrology and hydraulics section of the RI/FS Reportneed to contain the kind of detailed information listed above to adequately develop and evaluatecleanup alternatives. The DOT contamination is located in the St. Louis Impoundment which isimmediately up-river of the St. Louis dam and down-river of the Alma dam. Water flows andvolumes in this part of the river are regulated by the dams. EPA evaluated all practical cleanupalternatives. Capping sediments in place was found to be less favorable than sediment removalfor several reasons, (1) the water level in the impoundment in large areas is very shallow and acap would produce "islands" in the Impoundment, (2) capping would require long-termoperations and maintenance, sediment removal would not, and (3) EPA could not be certain thatcapping in place would reduce the level of contamination in fish tissue, whereas removal ofsediment is a permanent solution. The 1986 flood was not mentioned specifically, but scouringof a cap was considered.
Comment #12-2: Section 2.2.1 - Source Nature and Extent - Sediment.
#12-2a: Table 2.2-2 does not appear to contain the 1996 sediment data as indicated sinceit is identical to Table 2.2-8 which is only 1997 sediment data.
Response: Thecommenter is correct, EPA did mistakenly leave the 1996 sediment data out ofTable 2.2-2. The corrected table is as follows:
Table 2.2-2: Total DOT Concentration per Depth Interval (1996/1997 Data)Depth Interval (in.)
0-6
6-30
30-54
54-112
Max. Cone, (ppm)
229
32,600
32,600
822
Avg. Cone, (ppm)
25.4
1300
1320
102
Min. Cone, (ppm)
1.3
0.2
0.42
0.42
#12-2b: The use of simple arithmetic averages, while a mathematically correct method tocompute means, do not present a realistic description of the central tendency and
17
distribution of the DOT concentration data due to the high skewed distribution thatincludes occasional extreme high values.
Response: EPA did not calculate "simple arithmetic averages," the ArcView/ Spatial Analystsoftware that was used computed the mean for each interval. The use of the word "averageconcentration" in the RI/FS Report data tables 2.2-2, 2.2-4, 2.2-5, 2.2-6 and 2.2-8 was incorrectlyused.
#12-2c: In the 1997 data set, maximum, average and minimum DDT concentrationspresented are all influenced by the unconventional method of handling non-detects. Thenon-detects were reported as having been detected at the detection limit of 15 ppm. Inaddition, the RI/FS report states that 7 samples we:e greater than 1,000 ppm when thereference table indicates only 5 samples from 3 cores were above 1,000 ppm.
Response: EPA did assign all sediment sample non-detects the value of the detection limit, andalthough this may be a conservative approach EPA does not consider it to be "unconventional."The reference table in Appendix A of the RI/FS Report is correct with 5 samples that weregreater than 1,000 ppm. The 5 samples were used for the data analysis. The data from the labshowed 7 samples with levels greater than 1,000 ppm. Two of the 7 samples were matrix spikesand therefore were not included in the data analysis.
#12-2d: Part of the possible increase in average DDT concentrations observed in 1997 isdue to the inclusion of total DDT isomers in the 1997 chemical analysis that were notpreviously included in the analysis of total DDT concentrations. Based on the data tablepresented, inclusion of the para-para isomers may increase total DDT concentrations byas much as 4 times.
Response: EPA's analysis for para-para isomers of total DDT gives a more complete total DDTconcentration in sediments and indicates the concentrations were underestimated in the past.This also occurred with the fish tissue data. The 1997 fish tissue data was also analyzed forortho-para DDD and DDT, which had not previously been analyzed for. This increased fishtissue concentrations since significant levels of ortho-para DDD were found in fish tissue. Thisadditional chemical analysis shows that the concentrations in fish tissues were alsounderestimated in the past. Analysis of para-para-DDTr only (pp-DDT + pp-DDE + pp-DDD)carp fillet data show that the increase in fish DDTr content over the last decade is not an artifactof later inclusion of ortho-para-DDTr analytes. Lipid-normalized pp-DDTr concentrations haveincreased 50 to 70 % between the 1980s and the late 1990s below and above the St. Louis dam,respectively (Table A and Figure A).
The observed increase in pp-DDTr concentration in carp fillets also is not an artifact of changinglipid content over the years as shown in Figure B. Impoundment carp had much greater lipid-normalized pp-DDTr concentration in 1995 than in 1989 while lipid content remained constantover the same interval. The downstream carp also do not exhibit a linear relationship between
18
lipid-normalized pp-DDTr concentration and iipid content. These data confirm there has been asubstantial increase in the bioavailability of pp-DDTr in the Pine River system over the lastdecade. The expected natural attenuation cf the impoundment: sediment DDTr has not occurred.
19
Figure A. Lipid-normalized Total pp-DDTr Concentration in Carp Fillets Collected from thePine River Above and Below the St. Louis Dam, MI.
~ 800a.a.r 700a£.600O.15o 500•a<D
400
? 300
200
~J
1985 1989 1994 1995 1997Year
Locationx impoundmento below dam
Table A. Skin-off Fillet Carp Samples, Pine River, Ml, 1985-1997, Above and Below the St.Louis Dam, Michigan.Year
1*85
1989
1994
1995
1997
1997
Location
below dam
impoundment
below dam
impoundment
below dam
impoundment
Numberof fish
8
10
10
10
12
8
Weight
mean
kg1.209
1.149
2.001
1.789
2.819
3.099
Lipid Cont
mean
%
4.45
2.08
5.53
2.09
6.56
3.15
pp-DDTR Cone
whole tissue
ppm
9.6
9.5
22
15.1
21.3
24.6
lipid-normalized
ppm
2U.73
456.73
397.83
722.49
324.70
780.95
20
Figure B. Lipid-normalized Total pp-DDTr Concentrations versus Lipid Content in Carp Fillets
800
g; 700
QG6.Q.
N
oc
600
600
400
300
200
1997
1995
1989
1985
1994
1997 -
3 4 5 6 7Carp fillet lipid content (%)
Location
x impoundmento below dam
Collected from the Pine River Above and Below the St. Louis Dam, MI.
21
#12-2e: Table 2.2-3 in the Rl/FS report indicates that more than 90% of the DOT mass iscontained in under 10% of the sediment volume, yet there is no analysis for any actionlevel other than 5 ppm and 1 ppm.
Response: Based on data analysis, EPA concluded that the area, volume, and mass for total DOTlevels greater than 5 and for total DOT levels greater than 10 ppm both contained the sameboundary areas. A clean-up action level of Ippm did not pose a significant decrease in the riskwhen compared to an action level of 5 ppm, but was almost double in volume. Therefore u.iaction level of 5 ppm was decided upon because of its greatest reduction in risk that would allowfor DOT in the river to be at a safe level for ecological and human health.
#12-2f: On page 15, the report states that volumes were calculated using the results ofthe Inverse Distance Weighting (IDW) technique for sediment depths; however, later onpage 80 a uniform depth of 5 feet is used, yet the same volumes are obtained.
Response: Page 15 of the RI/FS Report is correct in stating that the volumes were calculatedfrom sediment depths using the results of the IDW interpolation method. The use of a uniformdepth of 5 ft on page 80 was to give the reader a relative number that could be used by the readerto quickly calculate an estimated volume.
#12-2g: On page 16 of the RI/FS Report it states that "interpolations are able to simulatea more realistic distribution of concentration." This gives a misleading sense of theaccuracy of the interpolations because (1) there is no alternative to which theinterpolation is being compared; (2) it doesn't take into account the effect of a fewextreme values in a sparse sampling network in overestimating the extent of highconcentration zones; (3) the use of only maximum concentration for each core willoverestimate concentrations of environmental relevance, especially when the maximumfor some cores were deeper than 30 inches and (4) because of the way non-detects werehandled the interpolation technique will overestimate concentrations at the lower end ofthe concentration ranges presented.
Response. Regarding the comment on page 16 of the RI/FS Report; in order to determine thebest spatial representation of the data, determine surface-area weighted concentrations for riskassessment purposes, and calculate volume and mass of contaminants in the impoundment, threepossible methods existed: depth-weighted arithmetic average, thiessen polygons, and inversedistance weighted (IDW).
Depth-weighted averaging was removed from consideration because of the ability to do spatialinterpolations, and more accurately represent the extent of contamination. Depth-weighting canbe useful in small, concentrated areas of interest, but may artificially inflate concentrations whenfew samples are available over a large area, and large values skew the results.
22
The thiessen polygon method involves constructing polygons by connecting the midpoints oflines between each sample point. Sample concentrations are then multiplied by the areaassociated with each polygon, and divided by the total area of interest. In this manner, there isno gradation of concentration between sample points., only blocks of concentration.
In comparison, inverse distance weighting allows creation of concentration contours, and a morevisual picture of the concentration range. IDW was the preferred method due to its ability tosimulate a more realistic distribution of concentration. Samples with high concentrations onlyinfluenced cells in a close proximity, whereas with thiessen polygons, actual concentrations mustbe used for a large surrounding block. In IDW, a grid is created of equal size cells, each cell isthen assigned a value according to a weighted formula of nearest neighbor sample points (i.e., thecloser a sample point, the more its concentration influences the value of the grid cell). Tocalculate a surface-area weighted concentration, the concentration of each grid cell was summed,and divided by the total area.
IDW interpolations, however, are considered more realistic due to the localized nature ofcontamination. When large gaps exist between sample points, and a single value is assigned tothat area, average concentrations will be increased. Therefore the IDW method was preferred.
Maximum Concentrations were taken per sediment core depth interval( 0-6 in, 6-30 in, 30- 54 in,and 54 - 112 in) not per sediment core. The data were queried per interval. The maximums foreach separate interval were taken. Each interval layer was then interpolated separately. Allintervals were interpolated by its maximum value.
#12-2h: On page 24 of the RI/FS Report, if the sediments are remaining in place onewould not expect significant change in the concentrations as a whole and especially atdepth where maximum concentrations are found. The declining trend in DDTconcentrations at the surface is not discussed in the RI/FS.
Response: Rivers are dynamic and unpredictable systems. It is feasible that the highlycontaminated sediments that are currently at depth may someday become exposed to the surface.The declining trend in DDT concentrations at the surface is not highlighted in the RI/FS Reportbecause it obviously has not positively impacted fish tissue concentrations. The fish tissue datashows that the decrease in sediment surface concentrations is not resulting in any decrease ofrisk to human health or the environment.
Comment #12-3: Section 2.2.1 - Source Nature and Extent - Fish Data. EPA's reportidentified "trends" and "significant increases" in concentrations in fish tissue without anystatistical evaluation. Most of the increase of total DDT levels reported in 1997 fishsamples are actually due to a change in analytical method, not necessarily changes in fishtissue concentration. No statistical adjustment is made for the potential influence ofincreasing weight and lipid concentrations. Other factors such as gender, season ofcollection and location of collection could influence DDT concentrations in fish. The
23
shortage of benthic organisms in the St. Louis Impoundment would account tor thedisturbing trend that carp collected in the Impoundment have less than 50% the fat offishcollected below the dam.
Response: The commenter does not indicate what type of "statistical evaluation" it believes isappropriate. EPA does not agree that the increased levels of total DOT in 1997 fish tissue is dueto a change in analytical method. EPA did look at lipid normalized fish tissue data, see EPA'sresponse to Comment #12-5. EPA agrees that a lack of benthic organisms in the Impoundmentmay be one explanation for the decreased % fat in fish collected from the Impoundment.
Comment #12-4: Section 2.3 Streamlined Risk Assessment.
#12-4a: The use of maximum detected sediment concentrations as exposure pointconcentrations is extremely conservative to the point of being unrealistic, and isinconsistent with EPA 1992 guidance.
Response: In addition to using the maximum concentrations for the RME scenario, the averagewas used in the central tendency scenario (see Table 2.3-4, page 40 in the RI/FS Report). Themaximum was used because in some cases the highest detected levels were near the river's edgeor near homes, therefore the maximum might very well be what people are exposed to. Becauseof the site-specific circumstances, it is not unbelievable that such exposure will actually occur.
#12-4b: The risk assessment does not factor in the effects of the hot spot removal.
Response: EPA had completed the baseline risk assessment and based on the results of thebaseline risk assessment determined that a time-critical removal action would be appropriate forsediments in excess of 3,000 ppm total DOT. Therefore the decision to conduct the time-criticalremoval was an outgrowth of the baseline risk assessment, not the other way around. In addition,as of February, 1999 no sediments have yet been removed from the Pine River, therefore itwould not be appropriate to assume this work is completed when it is not.
#12-4c: The approach used for dermal contact is inconsistent with that used at othersites. EPA guidance (1994) suggest that limited dermal contact be considered. At othersites, in the absence of beaches, dermal contact has been eliminated as an exposurepathway.
Response: Site-specific information is always an important consideration, and in this case it isvery relevant to determining whether dermal contact should be evaluated. Many homes borderthe river with no barriers at all; several homes have their backyard abutting the water (no fence,just grass and then water). Therefore, river overflow and access is obvious. A toy could bethrown from a backyard and land in the river. Therefore, EPA felt the exposure was quite realand important to assess.
24
#12-4d: The equation on page 34 is for estimating incidental ingestion of sediment, yetthe equation uses fish data as the concentration term.
Response: The commenter is correct, there is a typo with the equation on page 34 of the Rl/FSReport. It is the incidental ingestion equation, but the commenter was incorrect in stating thatCS is concentration in fish, CS is the concentration in soil. This equation should be taken outsince incidental ingestion was not done. The fish ingestion equation should be put in instead,and the CS term in the equation should be CF and ihe conversion factor removed. The correctedsection is:
Fish Ingestion Exposure EquationEquation for estimating exposure intake to contaminants due to incidental ingestion of
chemicals (USEPA RAGS, 1989).
Exposure = CF X IR X FI X EF X EDBWXAT
where:
CF concentration in fishIR ingestion rateFI fraction ingestedEF exposure frequencyED exposure durationBW body weightAT averaging time
#12-4e: The report says it likely underestimates risk, however the conservative exposurevariables and maximum concentrations used will tend to overestimate risk.
Response: EPA relied on the reasonable maximum exposure (RME) scenario for estimatingrisk, which most likely does not overestimate the risk at the site.
Comment #12-5: Section 2.4- Ecological Risk Assessment.
#12-5a: The use of whole-body carp concentrations to evaluate potential intake forherons is unrealistic. According to EPA (1993), 95% offish consumed by great blueherons in a Wisconsin population were less than 25 cm in length.
25
Response: Uncertainty associated with the size of the carp is discussed in the ecological riskassessment (ERA). This may result in an overestimation of risk to heron, however, it is the onlyavailable data for performing an ERA at the site. The possible overestimation of risk due to useof contaminant data from over-sized fish is probably more than compensated for by the use of thekestrel toxicity reference value (TRY) instead of the TRY for brown pelican, which is an order ofmagnitude more sensitive to the effects of DOT than kestrels. The difference in DOT levels inlarge and small carp is unlikely to approach an order of magnitude.
#12-5b: BSAF is used to ma^e the connection between sediment concentration and risk,but there are several problems with using it: (I) no explanation for the different BSAFbetween Impoundment carp and downriver carp; (2) it is unclear what sediment depthsare being used in the BSAF calculation; and (3) what values for fish DDT concentrationsand sediment DDT concentrations were used to derive the BSAF coefficient of 0.207presented on page 68; and (4) vastly different BSAFs would be computed using fish dataand sediment data from the Impoundment for different time periods.
Response: 1) Impoundment and downriver BSAF differences. Downriver carp were notassessed in the ERA because whole-body analyses were not available. However, the biota-sediment accumulation factors (BSAF) for the impoundment and downriver carp would not beexpected to be the same because of differences in exposure routes between the two locations.Fish in the impoundment are exposed both to direct absorption of dissolved DDTr in the watercolumn and to food chain uptake - probably through partitioning of DDTr from sediments todetritus (the remains of plants). The downriver reaches have few depositional areas. This meansthat downriver sediments are not a significant source of exposure to downriver fish. Instead, thedownriver fish probably receive the majority of their exposure from the water column. Theexposure route for downriver fish is the following: impoundment sediment DDTr - partitioningto impoundment surface water - transport of dissolved DDTr downstream - direct uptake fromsurface water by downriver fish. The significant relationship for downriver fish contaminantlevels is therefore with downriver surface water DDTr concentrations, not with downriversediment concentrations. Support for this hypothesis is the close agreement between the back-calculated downriver surface water DDTr concentration (0.2 ppb), based on downriver lipid-normalized carp fillet data and assumed water column only exposure, and the measured surfacewater DDTr concentration (0.3 ppb) below the St. Louis dam (see 4 below for calculations).
2) BSAF and sediment depth. The BSAF was calculated on the basis of the surface areaweighted total DDTr (op- and pp-DDTr) concentrations in surficial sediments (0 to 6 inchesdepth).
3) Derivation of BSAF (fish and sediment values). The BSAF value of 0.207 used in the ERAfor whole carp is based on a surface area weighted TOC-normalized total DDTr concentration of5032.9 ppm (dw) in surficial sediments throughout the impoundment, and a lipid-normalizedtotal DDTr concentration of 1040.1 ppm (ww) in whole carp.
26
4) Differences in BSAF over time is an important point. The lipid-normalized carp fillet datashow that DDTr has become increasingly more bioavailable over time (Table 1 and Figure 1). Ifthis trend continues, the BSAFs based on 1997 data will likely underestimate futureaccumulation in river biota. This provides strong evidence that the former approach to allow thecontaminated sediments in the St. Louis Impoundment to naturally attenuate, presumably bydeposition of clean sediments over the contaminated ones, was inadequate for this site.
There are two forms of DDTr: para, para- (pp-DDTr) and ortho, para- (op-DDTr). The former(pp-DDTr) are more toxic and persistent than the latter (op-DDTr), and comprise the majority ofthe DDTr in the river biota. As part of an evaluation of contaminant trends in the Pine River, pp-DDT, pp-DDD and pp-DDE were summed to determine pp-DDTr concentrations in carp filletscollected downstream of the St. Louis dam (op-DDTr was not included because it was notanalyzed in samples collected prior to 1997 - the sum of the op- and pp-DDTr is 25 - 38 % higherthan pp-DDTr in 1997 downstream and impoundment carp fillet samples, respectively).
The trends in bioavailability are best seen in the downstream data because the exposures arepredominantly through surface water absorption (see 1 above). The concentration of pp-DDTr is21.3 ppm in downstream carp fillets in 1997. Since DDTr is highly lipophilic ("fat loving"),virtually all of it is stored in fat tissue. For this reason, DDTr data are converted to a lipid-normalized basis, that is, the concentration in fillets is divided by the fillet fat content (6.56 % in1997 downstream carp fillets) to determine the concentration in fat tissue only. The 1997downstream carp lipid-normalized DDTr concentration is 324.7 ppm (21.3 -^ 0.0656). Under theassumption that the predominant route of exposure to downstream fish is direct absorption fromthe surface water (partitioning from the water column), the relation between surface water andlipid-normalized concentrations can be estimated from the octanol-water partitioning coefficient(Kow), which is the ratio of the concentration of a chemical dissolved in octanol divided by theconcentration in water. The log Kovv for DOT is 6.19 (ATSDR 1994). Log Kovv is presented as abase 10 logarithm, which means it is an exponent of 10 (i.e., 10 6 I 9 = 1,548,817). In this case,DDT occurs in octanol (a surrogate for fat) in concentrations over 1.5 million times higher thanin water. Divide the lipid-normalized DDTr concentration by Kow to estimate the surface waterDDTr concentration (Duursma and Carroll 1996), which, for 1997, equals 0.00021 ppm,equivalent to 0.21 ppb. This approach depends in part on how reliably octanol serves as asurrogate for lipids. A review of several studies showed that for persistent organic compounds,"oc'anol provides a reasonable representation of biota lipid for compounds with log Kow valuesbetv.een 2 and 6.^ (Connell 1998). The calculated downstream surface water DDTrconcentration (0.21 ppb in 1997, 0.27 ppb in 1994) is close to the measured concentration belowthe dam (0.3 ppb sampled 8/4/98). In contrast, the calculated downstream surface waterconcentration was 0.14 ppb in 1985. Between the mid-eighties and the mid-nineties, downstreamsurface water pp-DDTr concentrations increased by 50 to 86 %. Since there are no identifiedlocal downriver sources, this indicates that the rate of release of DDTr from the impoundmentsediments to the surface water has increased over the last decade.
The calculated impoundment surface water pp-DDTr concentrations range from 0.3 ppb in 1989
27
to 0.5 ppb in 1997. Both are greater than the measured surface water concentration at the MillStreet Bridge, 0.21 ppb, sampled 8/4/98. The difference is probably due to the additional foodchain exposure to fish in the impoundment. However, the data show that the combined surfacewater and food chain exposures to carp in the impoundment have increased by 67 % between1989 and 1997, consistent with the trend in downriver data.
28
Figure 1. Lipid-normalized Total pp-DDTr Cor .entration in Carp Fillets Collected from thePine River Above and Below the St. Louis Dam, MI.
Table 1. Skin-off Fillet Carp Samples, Pine River, MT, 1985-1997, Above and Below theSt. Louis Dam.
~ 800Q.Q.r rooQ
i.600Q.
o 500T30)
'% 400EOc 300
2001985 1989 1994 1995 1997
Year
Locationx impoundmento below dam
Year
Location Carp Body weight Lipid Content pp-DDTr Concentration
198519891994199519971997
below damimpoundmentbelow damimpoundmentbelow damimpoundment
number
8101010128
meankg
1.2091.1492.0011.7892.8193.099
mean%
4.452.085.532.096,563.15
whole filletppm
9.69.522
15.121.324.6
lipid-normalizedppm
215.73456.73397.83722.49324.70780.95
Literature Cited:
29
ATSDR. 1994. Toxicological Profile for 4,4'-DDT, 4,4'-DDE, 4,4'-DDD (Update). U.S. Dept.Health & Human Services, Agency for Toxic Substances and Disease Registry. TP-93/05. 166 p.
Connell, D. 1998. Bioaccumulation of chemicals by aquatic organisms. In G. Schuurmann and R.Markert (eds.). 1998. Ecotoxicology: Ecological Fundamentals, Chemical Exposure, andBiological Effects. Wiley, New York. pp. 439-450.
Duursma, E. and J. Carroll. 1996. Environmental Compartments: Equilibria and Assessment ofProcesses between Air, Water, Sediments and Biota. Springer, New York. 277 p.
Edwards, R., H. Egan, M. Learner and P. Mans. 1964. The control of chironomid larvae inponds, using TDE (ODD). J Appl Ecol 1: 97-117.
Matsumura, F. 1985. Toxicology of Insecticides, 2nd ed. Plenum Press, New York. 598 p.
NIOSH. 1994. NlOSH Pocket Guide to Chemical Hazards. National Institute for OccupationalSafety and Health. U.S. Gov. Printing Office, Washington, D.C.
Polder, M. and E. Hulzebos and D. Jager. 1995. Validation of models on uptake of organicchemicals by plant roots. Environ Toxicol Chem 14: 1615-1623.
Webber, E., D. Bayne and W. Seesock. 1989. DOT contamination of benthic macro invertebratesand sediments from tributaries of Wheeler Reservoir, Alabama. Arch Eviron Contam Toxicot 18:728-733.
Comment #12-6: Clean-up Goals.
#12-6a: There is an error in the volume break point analysis. The removal of materialfrom 5 to 1 ppm interval results in removal of 254,300 cy and 4,400 Ib. (0.0173 Ib/cy or17 ppm). In other words the additional material removed between the 5 ppm and 1 ppmDDT concentrations has an average concentration of 17 ppm according to the data aspresented in the table.
Response: EPA does not understand the commenters concern with the volume break pointanalysis (Table - 5-1).
#12-6b: The basis for the post-remedial fish DDT concentration calculation is withoutany supporting documentation. The assumed residual sediment concentrations used incomputing fish concentrations and the risks are unexplained. Is the assumption beingmade that a concentration of 0 ppm can be achieved by removal?
Response: Post clean-up goals were calculated using estimated post-clean-up sedimentconcentrations:
30
• 5 ppm clean-up goal with an estimated post clean-up goal of 1.5 ppm concentration insediment
• 1 ppm clean-up goal with an estimated post clean-up goal of 0.75 ppm concentration insediment
Using the estimated post clean-up goal sediment concentrations, we were able to back calculatefor a post-remedial fish concentration as shown in the Table 1 for clean-up goal of 1 ppm andTable 3 for clean-up goal of 5 ppm. Once post-remedial fish concentrations were established, wewere then able to calculate the post remedial risk for the two clean-up goals of Ippm and 5 ppmas shown in Tables 2 and 4 respectively.
Table 1Post Clean-Up Goal / Risk AssessmentPine River Supertund Site
Clean-Up Goal = 1 ppm
Concentration In Fish Post- Cleanup -> CF=(CSxBSAFxlipid) /( toc)
Species CS toc BSAF lipidsmallmouth bass 0.74 1 .23 0. 1 59 1 .66smallmouth bass-75 ucl 0.89 1.25 0.166 3.86smallmouth bass-95 ucl 1.12 1 .29 0. 1 79 3 .96smallmouth bass- whole 0.74 1.23 0.171 3.80carp 0.74 1.23 0.296 3.15carp- 75 ucl 0.89 1.25 0.357 3.91carp- 95 ucl 1.12 1.29 0.454 5.18carp- whole 0.74 1.23 0241 6.17carp- whole- 95 ucl 1.12 1.29 0.385 6.85
t^ic
£:3:;:£
;ji
I
|[
11
Ml|I Nl iI IllP•Aft
'll11I NIIllli(8*m
31
Table 2
CFBWATIRFlABEFEDslopeRISK
carp CancerCFBWAT
IR
FI
AB
EF
ED
slope
RISK
0.160170
255500.020.25
1365
90.34
5.00E-07
sport
0.5671
7025550
0.02
0.25
1
3659
0.34
1.77E-06
lllllBi
iiill!iiiiiii80.4583
70255500.075
0.51
3659
0.34
1.07E-05
middle aui
0.996770
25550
0.0750.5
I
365<>
0.34
2.33E-05
0.632270
255500.13
11
36530
0.341.71E-04
2.033970
25550
0.13
1
1
365
30
0.34
5.50E-04
32
Post Clean-up Goal/ Risk AssessmentPine River Superfund Site
Clean-Up Goal = 5 ppmConcentration In Fish Post Clean-up CF-(CS x BSAF x lipid) /( toe)
Speciessmallmouth basssmallmouth bass-75 uclsmallmouth bass-95 uclsmallmouth bass- wholecarpcarp- 75 uclcarp- 95 uclcarp- wholeearn- whole- 95 ucl
CS toe1.461.561.721.461.461.561.721.461.72
1.231.251.291.231.^31.251.291.231.29
BSAF0.1590.1660.1790.1710.2960.3570.4540.2410754
lipid1.661 C£j. SO
3.963.803.153.915.186.177.94
i$ifc%is ij;S:jS:-:S;Hs;i:
-H-w-iSiiMM§MS3B?£SJSgK:
. jilt•mil
33
Table 4
POST CLEAN-UP RISK CLEAN-UP GOAL OF 5 ppm
smallmouth bass CancerCFBWATIRFlABEFEDslope[RISK
carp CancerCFBWATIRFlABEFEDslopeRISK
sport031
7025550
0.020.25
1365
90.34
9.79E-07
sport1.11
7025550
0.020.25
1365
90.34
3.47E-06
middle subsis080
70255500.075
051
3659
0.34
1.87E-05
middle subsis1.74
70
255500.075
0.51
3659
0 34
4 07E-05
09470
255500.13
11
36530
0.342.54E-04
3.1270
255500.13
11
36530
0348.45E-04
#12-6c: Also missing are post-remedial risks for 10 ppm and 100 ppm cleanup levels.This would make it possible to compare a risk-based break point.
Response: Post remedial risk associated with a cleanup level of 100 ppm was not assessedbecause it exceeds me calculated range of ecologically protective sediment concentrations by 12 orders of magnitude. A 10 ppm cleanup level is just at the high range of the calculatedecologically protective sediment concentrations. Serious consideration of this cleanup levelwould require a close second look at the uncertainties in the ecological risk assessment (ERA)that result in possible underestimations of risk - in particular, the lack of consideration of thejoint effects of simultaneous exposure to DDE, DDT and DDD (the protective levels calculatedin the ERA are based on the effects of DDE alone) and the selection of kestrel toxicity datainstead of" the much more sensitive brown pelican data. These uncertainties are of lessersignificance for cleanup levels below the upper margin of the calculated protective sediment
34
to
levels.
#12-6d: Also unexplained is why postulated rates of decrease differ for smallmouth bass(93.6%) and carp (96%) using the same removal scenario.
Response: EPA considers the difference between 94% rate of decrease and 96% rate of decreaseto be so similar as to be the same.
Comment #12-7: Identification and Screening of Remedial Technologies.
#12-7a: The FS only assesses two alternatives other than the No Action alternative anddoesn't include in-situ immobilization or natural recovery. The report demonstrates nosignificant difference in understanding of the system hydraulically, geotechnically andphysiochemically from 1982 when the No Action alternative was selected for Pine Riversediments.
Response: EPA considered all viable cleanup options, one approach to in-situ immobilization iscapping and EPA did evaluate this alternative. Another approach to in-situ immobilization is in-situ solidification which has been demonstrated on a very limited scale (Mannitowoc, WI)without success. See The ARCS Remediation Guidance Document, October 1994, EPA 905-R94-003. Natural recovery is the alternative that was selected in 1982 for Pine River sediments.The 1998 RJ/FS report shows conclusively that natural recovery has not been effective atreducing contaminant levels in fish. Another possible reason natural recovery was selected in1982 was because dredging of contaminated sediment was not commonly done and thetechnology was not developed as it is today.
#12-7b: The hot spot removal dredging alternatives (2B and 3B) were screened out fornot being protective of human health and the environment, why?
Response: The hot spot dredging alternatives were screened out because the 5 ppm total DOTcleanup level could not be achieved by limiting dredging to hot spots and therefore EPA does notconsider these alternatives to be protective of human health or the environment. The 5 ppm totalDDT cleanup level is associated with fish tissue contaminant levels that would be safe forhumans and fish-eating birds to eat.
#12-7c: On page 95 of the RI/FS Report, Table 3-7, Number 7, there is no support forthe statement that in-situ containment "...may not reduce the bioavailability ofcontaminants in fish." It is unclear what basis was used to design the cap and how theneed to replenish every 5 years was determined. The ongoing deposition of cleanermaterial and a recent report of DDT biodegradability (Renner, 1998) enhances theargument for an in-situ containment alternative or no action remedy.
35
Response: The commenter is incorrect when they say there is no support for EPA's analysis thatcapping may not reduce the bioavailability of contaminants in fish. The Chicago office of theArmy Corps of Engineers (ACOE) verifies Lhat there are many unknowns with how effective acap will be in reducing bioavailability of contaminants to fish. Physical disturbances, diffusionand advection all can compromise the ability of a cap to perform adequately. In addition,capping is not a proven technology in shallow water, nationally caps have only been installed atdeep water sites, and none of these sites have post-monitoring data to show that they have beeneffective at reducing contaminant levels in fish. Dredging on the other hand has been shownthrough post-monitoring data at the Waukegan Harbor site in Illinois to have reducedcontaminant levels in fish such that a fish advisory was removed. At the Black River site in Ohicdredging resulted in reduced numbers of tumors in fish. Deposition of cleaner material and DDTbiodegradability are not effective at reducing contaminant levels in fish as evidenced by the fishdata which clearly show that although contaminant levels in the top 6 inches of sediments isdeclining, contamination in fish tissues is not.
Comment #12-8: Detailed Analysis of Alternatives.
#12-8a: The FS does not take into consideration the effect of the time-critical removalaction which will alter the goals and assumptions of the FS. To optimize the removalscenario, various post-RA DDT concentrations could be evaluated vs the expectedreduction in mass and anticipated residual concentrations. Nowhere in the RI/FS is theexpected residual concentration of DDT presented. If the uptake mechanism for fish isnot known, how can EPA predict that a cleanup level of 5 ppm will result in significantfish tissue concentration decrease?
Response: The removal action currently being undertaken is consistent with the remedial actionselected and does not alter the goals and assumptions made in the FS. The removal action willremove sediments that contain greater than 3,000 ppm total DDT. What will remain after theremoval action will still be significantly above the cleanup goal of 5 ppm total DDT, the levelEPA estimates will result in edible fish for both humans and fish-eating birds. The expectedresidual concentration after the removal action will still exceed 1,000 ppm total DDT in someareas of the middle basin. As of this time, no contaminated sediment has been removed from theSt. Louis Impoundment. EPA anticipates actual removal of the contaminated sediment to beginin late spring, 1999. EPA used the information from the 1997 sediment data and fish data to finda relationship between contaminant levels in sediment and contaminant levels in fish tissue.EPA then back calculated what the level in sediments would have to be to result in acceptablelevels in fish tissue, this assumes that the uptake mechanism for fish is the same before thedredging and after the dredging. This number was 5 ppm total DDT. The volume break pointanalysis shown in Table 2.5-1, page 80 of the RI/FS report shows that in order to reach a cleanupgoal of 10 ppm total DDT, 260,330 cubic yards of sediment would need to be removed. Toreach 5 ppm total DDT, EPA would only need to remove an additional 2,000 cubic yards. The 5ppm total DDT level is more protective than 10 ppm for a negligible increase in cost.
36
#12-8b: The potential need for polish dredging or capping residual materials afterdredging has not been included in cost estimates. Residual concentrations at the dredgedsurface could be higher than at the pre-dredged surface. Deposition of cleaner material toreduce these concentrations could take several years.
Response: Most of the sediment removal will be completed down to the bottom of the river,which is clay, so there will be no need for polish dredging or capping of residuals in place. The
'clay will be sampled and if it exceeds 5 ppm total DDT EPA will dredge some of the clay also.Therefore, residual concentrations at the dredged surface will not exceed the pre-dredged surface.
The sediment data shows that deposition of clean material over the contaminated material is notadequately occurring, therefore EPA does not intend to rely on clean material coveringcontaminated material, since this has proven not to be effective as evidenced by the fish tissuedata.
#12-8c: Increases in contaminant concentration in biota following dredging is notuncommon.
Response: Its true that sometimes there is a short-term increase of contaminants in biota, thishowever, does not negate the long-term good that will result from removing the contamination.
#12-8d: What about re-suspension? There doesn't appear to be enough detailedinformation about site specific conditions such as particle size distribution and the basisfor adding drying agent at a 10% dose to evaluate the alternatives effectively.
Response: Re-suspension is being addressed through engineering controls such as silt curtainsand coffer dams and turbidity monitoring. Treatability studies on solidification agents andparticle size distribution tests were being conducted concurrently with the RI/FS report. Theresults of the tests were not reported in the RI/FS, however they will be available for review bythe public.
#12-8e: It is unclear how EPA has determined that the DDT-containing sediment is not ahazardous waste.
Response: As set out in the RI/FS, the contaminated sediments are not considered to be listedhazardous waste under the Resource Conservation and Recovery Act (RCRA), 42 U.S.C. § 6901et seq., because the contamination occurred primarily from the direct discharge of DDT processwastewater to the Pine River. U.S. EPA and MDEQ concluded that the primary source of theDDT contamination in sediment was industrial waste water discharges of Michigan ChemicalCompany (now Velsicol Chemical Corp.) based on review of site history. Michigan ChemicalCompany manufactured DDT until the late 1950s, and discharged its manufacturing processwaste through numerous outfalls into the St. Louis impoundment and Pine River. Historicalanalytical data on the discharges from the outfalls showed that this manufacturing waste water
37 ' 7
contained DDT. The DDT-contaminated sediment that is the subject of EPA's response actionis in the area of Velsicol's outfalls. Under RCRA regulations at 40 CFR § 261.33(d) comment,process manufacturing waste that contains DDT is not a listed waste under § 261.33.
The contaminated sediments were also tested by the TCLP (Toxicity Characteristic LeachingProcedure) and determined not to be RCRA characteristic hazardous waste.
Comment #12-9: Cost Tables. Upon review of detailed costs, we believe that, overall,estimates are low. While this generalization does not necessarily apply to every remedialoption, it appears consistent with the removal alternatives.
Response: EPA disagrees. Acceptable variance in costs for a treatability study can vary between-30% to +50%. Velsicol neither provides any data to substantiate their belief that overall costsare estimated low nor do they define which cost elements would be included in a dollar/volumevalue. Thus no accurate comparisons or evaluations can be made based on Velsicol'sinferences.
Cost estimates are based on prevailing costs that U.S. EPA pays under Superfund, and equipmentcosts are based on quotes obtained from vendors contacted for this project.
Comment #12-10: It is worth noting that a recent (9/4/98) issue of Superfund Week indicatesthat a contract for $6M has been issued for the removal of 21,500 cy withremoval over 120 days. The unit costs for the removal are significantlyhigher and production rates lower than those presented in the RI/FS costestimates.
Response: The six million dollars include project setup/mobilization costs, removal anddisposal costs. Some of the planned set-up/mobilizatiou activities and associated equipment willremain on site for the second phase and results in a higher dollar value per yd3. These costs areusually spread over the entire project life and not just during the initial phase of the removalactivities. Since elevated levels of DDT-contaminated sediments are addressed in the initialphase of the removal activities, U.S. EPA intends to dispose of them at a subtitle C landfill, thusincurring higher disposal costs.
ALTERNATIVE 2A - Hydraulic Dredging, Dewatering, and Water Treatment
Comment 12-11: Without knowing the type and size of dredge being considered, theinferred production rate of 700 cy/day is optimistic based ondredging at other sites. Downtime is a significant factor and mustbe accounted for.
38
Response: A hydraulic dredge production rate of 700 yd3 per day was not presented in the RI/FSReport. EPA's proposed hydraulic dredging alternative assumed the use of two dredges toproduce an estimated flow rate of 6,500 gallons per minute (gpm) and a volume of 1,062 in-situyd3 of slurry per day. At U.S. EPA's Manistique Harbor site 500 yd3 of sediments per day weredredged from depths of 15 to 20 feet and transported by barge. The Velsicol project hasimproved upon equipment and material handling areas and would utilize 2 dredges to dredge at adepth of 2 to 3 feet and a slurry pipeline to transport sediments which will meet the givenproduction rate. The shallow depth and pipeline transport will increase rates of sedimentproduction.
The commenter indicates its comments are based on its dredging experience but fails to provide aproduction rate that it believe is realistic, nor does the c jmmenter give any specific informationabout the type of dredging that occurred or is occurring on these sites and if other factors playeda role in lower production rate, assuming the production rates were lower than EPA's estimate.
Regarding downtime, EPA estimated a 30% overall project downtime as part of the estimate.This project is planned to minimize overall downtime by using two hydraulic dredgessimultaneously, thereby not requiring a complete shut down of operations if one of the dredgesfails to operate. In addition, both the dewatering and water treatment units were designed tomeet the estimated daily production requirements. Multiple units of dewatering and watertreatment units will be available to minimize complete shutdown periods because of equipmentfailures. During short periods (1 to 2 hours) of complete shutdown, the dredged material wouldbe stored in settling tanks, inclined plate clarifiers, and the storage tank. When needed, thedewatering and water treatment units can be used during extended work shifts to treat storedmaterial and to maintain the anticipated daily production rate.
Comment 12-12: A major element that is missing from this estimate is the need fordebris and/or boulder removal.
Response: This comment is very generalized and does not account for site specific conditions.Based on our extensive sampling, the sediments do not indicate debris at levels of concern towarrant additional screening devices. The material dredged by hydraulic dredging would havesufficient settling time to settle out debris in the settling tanks. Cost estimates were projectedbased on information available at the time of their preparation.
Comment 12-13: It is unclear whether costs associated with moving silt curtains areincluded in the estimate.
Response: Costs associated with moving silt curtains is not listed separately but are assumed tobe part of day to day operation. The necessary equipment and crew is available on-site. An 11-hour dredging period out of a 12-hour work day is used for cost estimates. This 12-hour per daytime takes into account the movement of silt curtains and other work needed for securinghydraulic dredge areas. During the initial Superfund setup activities at Velsicol site, already
39
established silt curtains were moved in 25 to 30 minute intervals using barge and craneequipment. EPA does not anticipate the movement of silt curtains in these shallow waters to be atime problem.
Comment 12-14: The potential need for "polish dredging" or capping residualmaterial after dredging has.not been included in the cost estimate.
Response: The site is situated in shallow waters where the average zone of contamination iswithin the top 5 feet of the sediment column, and in many places this is the full extent ofsediment so all sediment will be removed down to the clay bottom of the river. The hydraulicdredge considered for this site has onboard global positioning system (GPS) to accurately locatethe dredge area and computerized systems to dredge to specific depths. Due to low flow regimesin the river, the potential for sediment migration is minimal and hence polish dredging was notconsidered unless abnormal concentration depths are encountered. Confirmation sampling willbe guided by GPS technologies and any polishing dredging to address abnormal concentrationswill be evaluated for specific areas based on this sampling.
Capping residual materials after dredging has not been considered for two reasons. Based onavailable information, the DDT-contaminated sediment area is well demarcated. This isexpected to help in the removal of DDT-contaminated sediments in excess of 10 parts per million(ppm) concentration with available technologies. At DDT concentrations of 5 ppm or less, theoverall adverse effects on aquatic animals is drastically reduced. The second reason being that atthese low residual concentrations, a very low potential for dissolved phase of DDT exists inwater.
Comment 12-15: The description of the alternative indicates environmentalmonitoring of treated water, downstream surface water, sedimentand ambient air during the 3 year project. Costs associated withthis monitoring program do not appear to be reflected in theestimate.
Response: Analytical costs for all kinds of monitoring are included in the fuel, utilities,supplies, analytical, miscellaneous materials and services subcategory of consumable materialsand services category of the cost estimate table.
Comment 12-16: Based upon our experience at other dredging sites it is likely that asignificant amount of time, effort and money will be spent to meetthe cleanup goal. Due to many factors residual sediments willremain after dredging and often at concentrations higher thanbefore dredging.
Response: The commenter is not clear on what the comment is and fails to cite specificexamples to substantiate their claim. It is true that meeting any kind of removal goal is difficult
40 - •"
when dealing with sediment contamination abatement. However, the hydraulic and mechanicaldredges (Liquid Flow Technologies and Cable Arm, Inc.) considered for this site will have GPSunits on them to accurately locate areas. Dredge units are computerized, and can traverse theclesired path with precision and could be programmed to overlap a certain percent of alreadydredged area. An added advantage to Velsicol site conditions is the presence of dams atupstream and downstream locations of the dredge area, that could be manipulated to createoptimal conditions during the removal alternative. Due to the shallow dredging depths, thedredge passes will be more accurate and the dredge units will encounter limited open waterdrifts. With all these capabilities, it is anticipated mat multiple passes will be minimized, oreliminated.
Comment 12-17: It is unclear what the "Construction Cost" line item includes.
Response: Construction costs include concrete pads, sheds, and other infrastructures needed forwater treatment plant, process and storage areas, and other areas of the project.
Comment 12-18: The text indicates that the remedial objectives "could be achievedin two or three construction seasons, but equipment costs are basedon only 13 to 14 months.
Response: Actual work would occur only during non-freezing and favorable climaticconditions. At the end of the work season, it is anticipated to demobilize equipment that is costprohibitive to keep on-site. Hence, such equipment costs are based on the duration of an actualwork period. Mobilization and demobilization costs are duly accounted for under mobilizationand demobilization costs for heavy equipment category of the hydraulic dredging alternative costestimate table.
ALTERNATIVE 3A - Mechanical Dredging, Dewatering, and Water Treatment
Comment 12-19: It is highly unlikely that the dredging production rate for thisalternative would be similar as for hydraulic dredging.
Response: It is true that typically a single mechanical dredge is less productive than a singlehydraulic dredge, but when more than one dredge is used simultaneously, this rationale wouldnot hold true. The cost estimates for this project, irrespective of the alternative selected, werebased on meeting preset production rates. Hence, for mechanical dredging alternative, twomechanical dredges would be used. The production rates of these dredges were obtained fromCable Arm, Inc., and a contingency factor of 12% was used to compensate for equipmentefficiency due to varying site conditions. The two mechanical dredges would produce acombined dredge rate of approximately 1,300 yd3 of in situ sediments per day.
Comment 12-20: Assuming that a 500 gpm water treatment system is adequate forhydraulic dredging, why would a similar capacity system be
41
needed for mechanical dredging?
Response; The commenter misunderstands EPA's treatment design. The water treatment isproposed to handle 500 gpm water for each carbon unit. The mechanical dredging proposeddesign would consist of 5 carbon units, the hydraulic proposed design would have 12 carbonunits.
Comment 12-21: Similarly, why would a dredge slurry pipeline be needed?
Response: The supernatant water/slurry accumulated in the initial container (i.e. barge or dumptruck) where the mechanical dredge first dumps the sediments will contain a mixture.ofsediments and water requiring treatment. The water and the suspended material would bepumped out through the dredge slurry pipeline for treatment.
Comment 12-22: It is unclear how the dredged material is to be dewatered.
Response: The mechanical dredge considered for this project from Cable Arm.inc., has specialfeatures that allow complete embedment of the bucket in the sediments to minimize waterlntakeand also has special sealing features to minimize water leaks. During EPA's treatability study, itwas observed that the site sediments dried up within 48 hours after the addition of a drying agentand exposure to atmospheric conditions. However, if addition of a drying agent becomes costprohibitive, an optional belt-filter press may have to be utilized. The cost of this equipment isnot included in mechanical dredging cost estimate table.
Comment 12-23: Comments 12-14 through 12-18 for Alternative 2A also apply toAlternative 3A.
Response: The responses provided for hydraulic dredging are also applicable to mechanicaldredging.
ALTERNATIVE 4 - Hydraulic Modification of Pine River, Excavation, Dewatering, andWater Treatment
Comment 12-24: The rationale for the removal production rate implied by estimateis unclear. Excavation in "the dry" is not necessarily faster thandredging, what is the basis for this assumption.
Response: The commenter did not specify production rates equipment types, hydraulic
42
modification procedures, river flow regimes, or other aspects of the project at the referenced sitesto justify their comment for this project. Two 4-yd3-DUcket capacity track excavators would beused to produce 1,732 yd3 of in situ sediments per 12-hour day.
Comment 12-25: It is not clear how the PORTADAMs are to be configured.
Response: This comment is more applicable to a full design document than a RI/FS report.Specific field configurations are usually detailed in a design document.
Comment 12-26: What about the potential need for "polish dredging?"
Response: This comment is not applicable for this alternative since excavation is not occurringunderwater and would not leave residual contamination or result in contamination migration.
Comment 12-27: What about costs for environmental monitoring?
Response: Analytical costs for all kinds of monitoring are included in the fuel, utilities,supplies, analytical, misc. materials and services subcategory of consumable materials andservices category of the cost estimate Table 4-5.
Comment 12-28: What does the "Construction Cost" line item include?
Response: Construction costs include concrete pads, sheds, and other infrastructures needed forwater treatment plant, process and storage areas, and other areas of the project.
Comment 12-29: The text indicates that the remedial objectives "could be achievedin two or three construction seasons
Response: Actual work would occur only during non-freezing and favorable climaticconditions. At the end of the work season, it is anticipated to demobilize equipment that is costprohibitive to keep on-site. Hence, such equipment costs are based on the duration of actualwork period. Mobilization and demobilization costs are duly accounted for under mobilizationand demobilization costs for heavy equipment category of the hydraulic dredging alternative costestimate table.
ALTERNATIVES 5 AND 6 - Disposal at a RCRA Subtitle D/C Landfill
Comment 12-30: The total tonnage of material appears significantly low. Assuming50% solids in-situ, an in-situ density of 80 pcf, and 60% solidsfrom the dewatering operations, we would expect a total of approx.260,000 tons. The mass of drying agents must also be added to thetotal tonnage.
43
Response: EPA assumed that one yd3 in situ volume of sediments translates approximately to0.65 tons of processed ex situ sediments for disposal based on our experience at the Manistique,Michigan dredging project. This translates to approximately 161,250 tons of sediments prior tothe addition of a drying agent. The added mass due to the addition of a drying agent at 10% isaccounted for in the disposal estimates. The combined total of sediments and drying agent is178,750 tons and is given under the Disposal of Nonhazardous Waste category of respectivedisposal alternative tables.
However, if a literature value (one unit volume of in situ sediments approximates to one-thirdunit volume of sediments) is used for the drying agent, the weight increases to 260,625 tons.EPA elected to base the estimate on our experience at Manistique, Michigan rather than on theliterature value.
Comment 12-31: It is unclear from the cost what landfill is being assumed (fortransport and tipping fee). Based on discussions with a subtitle Clandfill in MI, treatment and disposal for DOT sediment would beapprox. $300/cy. This is significantly greater than the $65/cy used.
Response: All costs are based on disposal as nonhazardous waste. Disposal costs at a subtitleD landfill are lower than at a subtitle C landfill. The quote obtained for this project is based onthe waste being classified as nonhazardous with no additional treatment being completed at thedisposal facility.
ALTERNATIVE 7 - In-Situ Capping
Comment 12-32: It is unclear whether the cost associated with moving the siltcurtain is included in the estimate.
Response: Costs associated with moving silt curtains is not listed separately but are assumed tobe part of day to day operation. The necessary equipment and crew is available on-site as seen inour response to this question under hydraulic dredging.
Comment 12-33: The description of the alternative indicates environmentalmonitoring will be conducted but the costs don't appear in theestimate.
Response: Analytical costs for all kinds of monitoring are included in the fuel, utilities,supplies, analytical, misc. materials and services subcategory of consumable materials andservices category of the cost estimate Table 4-7.
44
Comment 12-34: It is unclear what the "Construction Cost" line item includes.
Response: The construction costs for this alternative include material storage pads and materialloading pads.
45
APPENDIX D
U.S. BNVIRONMKNTAL PROTECTION AGBNCYREMEDIAL ACTION
ADMINISTRATIVE RECORDFOR
VELSICOL CHEMICAL COMPANY SITEST. LOUIS, ORATIOT COUNTY, MICHIGAN
DATE<-
00/00/00
05/31/55
07/00/57
00/00/75
10/00/80
04/00/82
12/27/82
AUTHOR
MDPH
MDNR
Courchaine,R., MichiganChemicalCorporation
Lincer, J.,MoteMarineLaboratory
Forba, R.,NationalEnforcementInvestigationsCenter
Forba, R.,NationalEnforcementInvestigationsCenter
U.S. DistrictCourt/EasternDistrict ofMichigan
ORIGINALFEBRUARY 10, 1999
UCIPTKNT
MichiganChemicalCorporation
U.S. EPA/Office ofEnforcement
U.S. EPA/Office ofEnforcement
Respondents
TITLE/DESCRIPTION PAGES
Toxic Substances Fact 3Sheet for DOT DDE andDDD
Report: A Biological 5Survey of the Pine RiverAbove Alma to M-30 toDetermine Effects ofPollution
Report of Survey: 11Waste Survey ConductedJuly 8-11, 1957 at theMichigan ChemicalCorporation Site
Journal Article: DDE- 13Induced Eggshell-Thinningin the American Kestrel:A Comparison of the FieldSituation and LaboratoryResults (J Appl Ecol;12: 781-793)
Report: Pine River 67Contamination Survey{June 2-6,1980)
Report: Summary of 30Pine River ReservoirSediment SamplingSurvey (November 20-22, 1981)
Consent Judgment re: 52the Velsicol ChemicalCompany Site
H&*. QAXB
8 06/00/89
APTHQR
USDHHS/USPHS/ATSDR
14 01/00/97 Blasland,Bouck &Lee, Inc.
15 10/01/97 MemphisEnv i ronmen t a1Center, Inc.
16 10/01/97
17 12/00/97
MemphisEnvironmentalCenter, Inc.
Conestoga-Rovers &Associates
U.S. EPA
9 09/30/93 USDHHS/USPHS/ U.S. EPAATSDR
10 00/00/95 International U.S. EPAJointCommission
11 09/00/95 USDHHS/USPHS/ATSDR
12 10/25/96 Sakowski, K., Reiner, BMDEQ U.S. EPA
13 00/00/97 International U.S. EPAJointCommission
U.S. EPA
U.S. EPA
U.S. EPA
U.S. EPA
VelBicol Chemical Remedial ARPag* 2
TITLR/DB3CRIPTIOH PAGES
Preliminary Health 7Assessment for theVelsicol ChemicalCorporation (St. LouisPlant Site)
Site .Review and Update 17for the Velsicol ChemicalSite
Paper; PAH Contaminated 5Sediment Remediation inthe Main Stein, BlackRiver (OH)
ATSDR Fact Sheet re: 2DDT, DDE and DDD
Report: Status of the 185St. Louis Impoundmentin the Vicinity of theFormer Velsicol ChemicalCompany w/ Cover Letter
Paper: PCS Contaminated 7Sediment Remediation inWaukegan Harbor (IL)
Report: Response to 15October 1996 MDEQ StaffReport on the Status ofthe St. Louis Impoundment
Containment System 863Assessment Report:Volume 1 of 2 (Text,Tables, Figures andAppendices A-H) [FINAL]
Containment System 452Assessment Report:Volume 2 of 2 (AppendixI, Results, ResultsAppendices A-F) [FINAL]
WorJc Plan: Post-Closure 31Cap Maintenance for theFormer Michigan ChemicalPlant Site
DAIS AOTHQR
18 02/20/98 Sargent, L.,MDNR
RBCIPIBMT
Hanshue, 5MDEQ
Velaicol Chemical Remedial ARPage 3
TITLB/DBSCRIPTIOIf PAG53
Memorandum re: St. 2Louis ImpoundmentSediment Remediation
19 06/09/98 Borries, S.,U.S. EPA
20 06/18/98 Muno, W.,U.S. EPA
Ullrich, D,U.S. EPA
McMann, T. ,Sidley &Austin andJ. Ray,NWI LandManagementInc.
Action Memorandum: 23Request for a Time-Critical Removal Actionat the Velsicol Chem-ical Company Site(PORTIONS OF THIS DOCU-MENT HAVE BEEN REDACTED)
Letter Forwarding the 25Attached June 16, 1998Unilateral AdministrativeOrder for Access to theVelsicol Chemical CompanySite
21 08/00/98 U.S. EPA/Region 5
U.S. EPA Streamlined RemedialInvestigation Report(Final) for theVeliscol Chemical Site
172
22
23
09/00/98
09/02/98
24 09/09/98
25 09/16/98
U.S. EPA/OPA
Morning Sun(Mt. Pleasant)
Public
Public
Morning Sun(Mt. PleasantJ
Public
Bay AreaReporting
U.S. EPA
Fact Sheet: ProposedPlan for the VelsicolChemical Superfund Site
U.S. EPA Public Noticere: Announcement of theSeptember 8 - October 8,1998 Public Comment PeriodConcerning Cleanup Alter-natives at the VelsicolChemical Site
U.S. EPA Public Noticere: Announcement of theSeptember 16, 1998 PublicMeeting Concerning theProposed Plan for theVelsicol Chemical Site
Transcript of theSeptember 16, 1998Public Meeting re: theVelsicol Chemical Super-fund Site
87
MO. DATS
26 10/00/9b
AUTHOR
ConcernedCitizens
RECIPIENT
U.S. EPA/
Velsicol Chemical Remedial ARPag* 4
TITLE/DESCRIPTION PAQBS
Six Letters re: PublicComments Received Septem-ber 3 - October 7, 1998Concerning the ProposedPlan for the VelsicolChemical Site
12
27 10/08/53 Phillips, J.,MemphisEnvironmentalCenter, Inc.
Hill, S.,U.S. EPA/OPA
28 10/23/98 Ecology andEnvironment,Inc.
U.S. EPA
29 10/23/98 Ecology andEnvironment,Inc.
U.S. EPA
Letter: MEC's Comments 36on the Proposed Planand Final StreamlinedRemedial Investigation;'Feasibility Study Reportfor the Velsicol ChemicalSite w/ Attachments
Draft Treatability 261Study Report: Volume1 of 2 (Text, Tables,Figures and AppendicesA-E) for the Velsicol/Pine River Site
Draft Treatability 622Study Report: Volume2 of 2 (AppendicesE-P) for the Velsicol/Pine River Site
DATE AUTHOR RECIPIENT
gOIDANCE ADDENDUM
V*l«icol Ch*mical R*m*dial ARPag* 5
TITLE/DESCRIPTION
THE FOLLOWING DOCUMENTS ARE INCORPORATED BY REFERENCEAND HAVE NOT BEEN COPIED FOR PHYSICAL INCLUSION
INTO THE ADMINISTRATIVE RECORD
00/00/78
00/00/80
Halbert, F. &S. Halbert
U.S. House ofRepresentatives
00/00/93 Lorenz
00/00/96 Colborn,et al.
Bitter Harvest (Michi-gan : Wm. B. EerdmansPublishing Co., 1978)
PBB (PolyJbromina tedBiphenyls) PollutionProblem in Michigan:Hearing Before theSubcommittee on WaterResources of the Com-mittee on Public Worksand Transportation{HR: 66th Congress,1st Session)
Journal Article: Con-taining the MichiganPBB Crisis, 1973-1992.-Testing the Environ-mental Policy Process(17 Environmental His-tory Review 49, Summer1993)
Our Stolen Future (NewYork: Penguin Books,1996)
V*lsicol Chemical Remedial ARPag* 5
DAIB AUTHOR RECIPIENT
GUIDANCE ADDENDUM
TITLE/DESCRIPTION PAGES
THB FOLLOWING DOCUMENTS ARE INCORPORATED BY REFERENCEAND HAVE NOT BEEN COPIED FOR PHYSICAL INCLUSION
INTO THB ADMINISTRATIVE RECORD
00/00/78
00/00/80
Halbert, F. &S. Halbert
U.S. House ofRepresentatives
00/00/93 Lorenz
00/00/96 Colborn, T,et al.
Bitter Harvest (Michi-gan: Win. B. EerdmansPublishing Co., 1978}
PBB (PolybrominatedBiphenylj) PollutionProblem in Michigan:Hearing Before theSubcommittee on Water-Resources of the Com-mittee on Public Worksand Transportation(HR: 66th Congress,1st Session)
Journal Article: Con-taining the MichiganPBB Crisis, 1973-1992:Testing the Environ-mental Policy Process(17 Environmental His-tory Review 49, Summer1993}
Our Stolen Future (NewYork: Penguin Books,1996)